1031 lines
369 KiB
Plaintext
1031 lines
369 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "08b51816-e857-4bd7-9bac-377198a61d4a",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"**Вставляем код из методички**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 14,
|
||
"id": "f078b767-66a4-4176-b88a-30774f78ca45",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" 0 1.00 1.00 1.00 12\n",
|
||
" 1 1.00 0.89 0.94 9\n",
|
||
" 2 0.90 1.00 0.95 9\n",
|
||
"\n",
|
||
" accuracy 0.97 30\n",
|
||
" macro avg 0.97 0.96 0.96 30\n",
|
||
"weighted avg 0.97 0.97 0.97 30\n",
|
||
"\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"from sklearn.datasets import load_iris\n",
|
||
"from sklearn.model_selection import train_test_split\n",
|
||
"from sklearn.neural_network import MLPClassifier\n",
|
||
"from sklearn.metrics import classification_report\n",
|
||
"\n",
|
||
"# Загрузка и разбиение данных\n",
|
||
"X, y = load_iris(return_X_y=True)\n",
|
||
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)\n",
|
||
"\n",
|
||
"# Модель MLP — многослойный перцептрон\n",
|
||
"clf = MLPClassifier(hidden_layer_sizes=(10,), activation='relu', max_iter=2500)\n",
|
||
"clf.fit(X_train, y_train)\n",
|
||
"\n",
|
||
"# Отчёт о точности\n",
|
||
"print(classification_report(y_test, clf.predict(X_test)))"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "4118b2a1-6118-4775-bf7b-ee2737a4c80c",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Классификация с использованием стохастического градиентного спуска**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "2d8a7b18-68d4-403f-94f6-ceed767c0362",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Цель задачи:**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "4fe5030f-2ead-4767-b153-56758e7604ed",
|
||
"metadata": {},
|
||
"source": [
|
||
"Продемонстрировать, как линейная модель классификации, обученная с помощью стохастического градиентного спуска, работает на разных типах данных.\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e6d4c96c-11c7-4dc1-b09c-bb9a08a89038",
|
||
"metadata": {},
|
||
"source": [
|
||
"`SGDClassifier` из модуля \"sklearn.linear_model\". Это эффективный метод для обучения линейных моделей на больших наборах данных."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "f25aad18-6913-4f7e-baa6-20675683beed",
|
||
"metadata": {},
|
||
"source": [
|
||
"**План анализа:**\n",
|
||
"1. Обучение модели на встроенном датасете Iris и make_blobs.\n",
|
||
"2. Повторение анализа с другим реальным датасетом\n",
|
||
"3. Сравнение результатов и визуализация решений."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 3,
|
||
"id": "ba96869d-7a19-41b3-bcc3-2156f3fdfb42",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHHCAYAAABHp6kXAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAAohxJREFUeJztnQd8E+Ubx39JuveglL1lyB6KgAxBwC2KCxXBicpGNgqCQlWQjSCigIqCoqh/F6AiCAIiyBLZexa6d5rk/p/nTS9NSselvezn+/lck7tc3nvvkia/PFMjSZIEhmEYhmEYD0fr6gkwDMMwDMOoAYsahmEYhmG8AhY1DMMwDMN4BSxqGIZhGIbxCljUMAzDMAzjFbCoYRiGYRjGK2BRwzAMwzCMV8CihmEYhmEYr4BFDcMwDMMwXgGLGsbnOX36NDQaDVasWGHX87p16yYWb+bKlSt46KGHEBsbK67R3Llz4ans2rULHTt2RGhoqDiXvXv3unpKeP3118VclEDvT9qX3q9q4AvvX8b3YFHDuBz5w1pegoKCUK1aNfTu3Rvz589HRkaGq6fos4wcORLr16/HhAkT8Mknn+COO+6AJ5Kfn4+HH34YycnJmDNnjjiX2rVrwx2ZMWMGvvnmG/gqJNqefvpp1K9fX3wWVKlSBV26dMGUKVOK3f9///sf7r33XsTHxyMgIAAxMTFi/3fffRfp6ek2+9apU8fyOaPVahEVFYXmzZvjhRdewM6dO510howj0XDvJ8YdRA19iE2bNg1169YVX0CXL1/G77//jo0bN6JWrVr47rvv0KJFC4ccn/4F8vLy4O/vD51Op/h5er1e3NIHqbdCXyi33347Pv30U3gyhw8fRpMmTfDBBx/gueeeg7tAlpqpU6eK96BMWFiYsI4VtRwajUbxvxEYGKjYulMaspWG/s/chePHj+Omm25CcHAwnnnmGSFCLl26hD179uCnn35Cbm6uZV+TyYRnn31WXCcSJn379kXNmjXFj6Dt27fj22+/FZa5X3/91fIcGi86OhqvvPKKWKd9//vvP3z55ZfiM4dE/OzZs11y7ow6+Kk0DsNUmDvvvBPt2rWzrJN14LfffsM999yD++67T3z40Ied2sjWIXvxVjFjMBjEFwadX2Jiovg16+nQeRCefC4kuO0R3Z4IWdEyMzOFa7CoJU1+DWXeeecdIWhIiJBVxlroDR8+XIihjz/++LpjVK9eHU8++aTNtrfffhuPP/64OP4NN9yAl156SfVzY5wDu58Yt6Z79+547bXXcObMmeusBfTrm37RkrmZRAkJIrLoFCU1NVV88NGvNPqVW6NGDTz11FO4du1aiTE19KuNrEe0Lz2natWquP/++23iGYqLSaAPXvr1SKZwmlPLli2xcuVKm33k482aNQtLly4VZnY6Bv1CpbiPsqBf6/Trnj586RgU73LrrbcKq1ZpcyMGDhworkNxc6F4GXku7733nthOFoRFixZZTPYEuXBGjx4tfh2TVSEiIkII0n379l13PPplTdaIhg0birnSdXzwwQdx4sQJyz4koOjYTZs2FfvQtRs0aBBSUlKgBBK+nTt3FrEyJFrodSIBbH3OXbt2FffJBUXnUVosiewO3bp1K4YNG4a4uDgxLs2JrHP0fqL3D/3ip2Xs2LE2lhayfNDzi1pAlMRu0eNZWVniPSNfc5p/eWJq6P/l5ptvRkhIiJgnuWQ2bNhQ4v50bpMnT0bbtm0RGRkpridd102bNl237+rVq8V+4eHh4vWn98K8efPseo8WB70v6H+uONdg5cqVLfezs7OFEKH3zMyZM4u1XNF7bdy4cVAC/VgilyR9lkyfPt3m9WQ8C7bUMG5P//79MXHiRPGB/Pzzz4tt//77Lzp16iR+dY0fP158AH/xxRfo06cPvvrqKzzwwANiP/rVRx/M9CVH5uw2bdoIMUPi5/z586hUqVKxxyRTNh1j6NChQgSQWKEP5LNnz9qIAmtycnLElyWZ0IcMGSJcaWTWpi8l+iKkX4/WfPbZZ8L8TV+W9KFMvzzpC//kyZPCFVYSJBISEhKEG4W+tChu4O+//xYm+p49e5brGi9fvlwIEIotIFFD14k+5Ona05j0JS5D86OYDxIIdI4UTPz+++8L4XDo0CERDyW7S8jKRub/xx57TJw/nS9dx4MHDwoBRdD5yy5IEhGnTp3CwoUL8c8//2Dbtm2lXotffvlFCKp69eqJ60KvwYIFC8R7g64HvVY0Pr1PKFaFxifxSMKpLOi1J/cbfTnv2LFDCFASN3/++adwidJ4P/74o/hSbdasmc01Ki90zeXXlV4LQr5O9kBzputB7hdy65LVjWJGSAD26tWr2OfQ+2jZsmXo16+f+D+j1+rDDz8UsW1//fUXWrVqJfaj14/26dGjhxAWBP1/0Wslv8fL+x4lMUOvKc2TftCUBAlO+p8ica2W9YoEOn1u0DnT+5gEE+OBUEwNw7iS5cuX088iadeuXSXuExkZKbVu3dqy3qNHD6l58+ZSbm6uZZvJZJI6duwo3XDDDZZtkydPFmN//fXX141J+xOnTp0S+9A8iJSUFLE+c+bMUufdtWtXscjMnTtXPO/TTz+1bNPr9VKHDh2ksLAwKT093eZ4sbGxUnJysmXfb7/9Vmz/3//+V+pxW7ZsKd199912zU1mwIABUu3atS3r8lwiIiKkxMTE6/anxwYPHmyzja650Wi02UbjBAYGStOmTbNs++ijj8TzZ8+eXeK1/+OPP8Q+q1atsnn8559/LnZ7UVq1aiVVrlxZSkpKsmzbt2+fpNVqpaeeesqybdOmTWK8L7/8UlL6fuzdu7dlngS9jhqNRnrxxRct2wwGg1SjRg2bay0fi26LXiPr9xkxZcoUsc2a0NBQ8TqVNC8apzSOHTsmzv+BBx647nWyPp+i7xE6l7y8PJv96X8hPj5eeuaZZyzbhg8fLt4vtH9F3qPFcfDgQSk4OFicJ722dKxvvvlGysrKstlv3rx5Yh96zBqa09WrV20W63Om935p85ozZ44Yl/4XGc+E3U+MR0C/ouQsKHJ/0C+5Rx55RGwjywstSUlJ4lflsWPHcOHCBbEvWW3IBSRbbqwpKdiSTNH0y5bcB0pdIAT9aqdf9vQrVoasDGQdIIvR5s2bbfZ/9NFHhVtAhixKsiWkNMhaQFYkOk+1IMsUuVmUQJYcyhyRrTF03en1adSokfglLkPXnixhZPEo6dqTJYtcHfTrXX4daSHXBo1ZnOtDhmImKPaCLGHkNpChgHIaj16PikBuROv3SPv27YVbgrbLkJWA3J5lvWbOhKxo5NIjV5L8OsmUFmBM5yLHidHz6f+M4qvo/KxfV3r/kYusNFdSed+jZB2h15RiXsjNRi4tsr6SZY2CvGXkrCZ6j1hz4MAB8T62Xuj9qRR5PM649FxY1DAeAYkC8t8T5N6hLxeKtSn6ASanfcpBheSjJ9eAPdCXNpnVKduCPkwpFoFcQxRnUxoU90MxBEW/SCjrRn7cGnJhWCMLnLKEFLkTyPROcSoUyzBmzBjs378fFYHcSEqhLzw5oJKuFQkXuvY0h7S0NMt+dO1J6Pj5lezlpi89eg7FSxR9Lek1Lxocao18PekYRaFrTuKIvnzLS9HXh8QXQRk2RbfbI37Vgq4bvSflhUSIfN3pPXjjjTfaPSbF8pAolONg6HX44YcfbF7Xl19+Wbz3yO1H8S/k1v35559Ve4/Sc8gNR68fPYfcfPQeInccuaYI+bOA3iPWNGjQQIgtWsh1ai/yePL4jOfBooZxeyj2hT5U6QNL/lIlyJ8uf4AVXeR9y8uIESNw9OhRERdAH/AkoOiLkuI81KKkWICyghRJZNEX10cffSQEG8VBUAwM3Zb1i5wsK8VhT1YZfcmMGjVKzIOCUamODV1z+pUtvzZKof1J0JT0OtKXo6so6fUpbrv1a2bvtS8vFL9CwbDyQvFYFYFeS7J6UQwPxZWQUKHXgGJbrF9Xer3ImkJxaZSVSNY0EjgDBgyw6z1aFnSdSRBRFuS6devEtlWrVonbxo0bi1uKzSpqaaESBLRQnJW9yONV9PODcR0cKMy4PfSrjSDXEiF/WJFrhz68SoM+oIt+8CmFnkv1LGghiwIFSlLqaEk1WyjIkX5Z0heAtbWGsrTkx9WC3C0UWEsL/bqkLxEKzpRrsJDVpziXSFFrUXlYu3YtbrvtNvHFZw39MrcOvKbrR8GplAlTUrAv7UO/vimw1950ffl6Hjly5LrH6JrTXCiA3NnIFje6HuW59kpr0FDWlXVqsnxcuqb0HqRgVzm4V+nrSv9bX3/9tc0ciit6R24qKnhHCx2LrDcULE7iXxYEZb1H7UEu9UAuR9lVSxYyysIi0VPUOloeaI4knsgSJ1tXGc+DLTWMW0OxM2+88YZwjzzxxBOWX4qUZUQfovKHnDVXr161iRWhVGP5l54Siwili1oX+ZK/KMgkTUX6SuKuu+4SboA1a9ZYtlFMAmXj0C9IOa24ohSNEaCx6YvEem40X/pit74WdB0oQ6Wi0C/ooteOYmPkOCbra08uBMpkKor8fIqLIgsGvcZFoWtXVBhYQ9YJ+tIml4n1fiRiKVOOXg9XQGKLrtGWLVtstlOavBJIiJV23jLkXpKtErRQHBJBMSj0JU9WrqKWs9KsgLIFynofEqVUyK609x8dSy6MKb8HlbxHi+OPP/4QIrgocnyU7GqkNHUSdfRaU/ZjcedlT1o2Zc2Ru4pceJMmTVKluCHjGthSw7gNFMNCX8T0ZUZpwiRoyPxNXxJk6rYukEe1U6juBZmnKf2UfmHSc+gDmNxVcs0U8uXTL1BKPybfP33w0wcXjbdkyRIRRFwUcjtRuip94dIXB/nzSRTR+JSaXBLk8yehRSb83bt3i3RiOjYJCarDopafnuZEoo7OhX4NU6osHYfSyGXoXKkyKlm3KLCVYlPofMlFVLR0vL1QmjZ9YdIvcEoZpuBMcgsUNfdTijMVPyNXFaUE069rinEhywz9sqd6MiT0KOWa3Hzk0qB0Y7LqkGWMhBIFilItopKgdGpyfXTo0EGcp5zSTb/iySrgCujY9H6jedCXIwnM77//vtT4IGvodaVrRK8fpceToKcgZaWQeKAvZhKKdM3JLUWxT1QDicaja13S60pWGgqqv/vuu0VqPb1n6P1mHbtClhb6HyK3FMXUkAWKzpUEpmzhUPIeLQ6KZaP/HZqzLJQoSJneRzQOuYVlSMxQKjm9B0jEkoim+VB8Ez2H3j/0A6hoYU0S37K1lc6LLFpyRWGyytL7kfFgXJ1+xTByqqq8BAQESFWqVJF69uwpUjflVOiinDhxQqTt0r7+/v5S9erVpXvuuUdau3atzX6U7jtkyBDxOI1NKbiUMnvt2rViU21pO6UxN27cWKTXUjp5+/btpS+++KLMtOkrV65ITz/9tFSpUiVxLEo7t07htT5ecSnjtJ3SfEvjzTfflG6++WYpKipKpL/SPKdPny7Sx62h1PJ69eqJeVB67Pr160tM6S4pfb2klO5XXnlFqlq1qjh+p06dpO3btxd7PbKzs6VJkyZJdevWFa8RvVYPPfSQeO2sWbp0qdS2bVsxXnh4uLhuY8eOlS5evCiVxS+//CLmQM+lVON7771XOnTokM0+5UnpLlpiQE6/pjRha+ia0vvEGtqnb9++UkhIiBQdHS0NGjRIpCsrSek+fPiw1KVLF0tqs5zerTSl2zqlnsogUKo9zYFem40bN1oeL/p6UerzjBkzxPuDnkPP/f777697z9D/V69evUQqPb23atWqJc7v0qVLdr9Hi7Jt2zbxfmvWrJn4v6P3DI0/cODA694zMuvWrZPuuusuKS4uTvLz8xPHvPXWW8V7OjU11WZfOg/5c4bS8+n90rRpU+n555+Xdu7cqei6Mu4N935iGIZhGMYr4JgahmEYhmG8AhY1DMMwDMN4BSxqGIZhGIbxCljUMAzDMAzjFbCoYRiGYRjGK2BRwzAMwzCMV+BTxfeouubFixdFETSuGMkwDMMwngFVn6Hu6VRAstS2GJKHkpCQIAooDR8+XPFzzp07Z1PkjRdeeOGFF154gccs9D1eGh5pqaFy31SOXi6jrRS5TP2Cn/5CcGiYg2bHMAyjApIR8RktoYEJiWG/waSt4vBDhkWtRYteS1B12X0YM70JelS/vm9XcRhNOui06nYhZxhrsvONePbbk2W2m/E4UUO9Oqix4QcffIA333zTrufKLicSNCFh6vThYRiGcQRa0yWEmUyQoENQeD1AY2446Sje3X8nvpmYiktLf8Gkt7fi3jqzqc1lsftuPNEXN1XfhKigZKutjp0fwxBlhY54XKDw4MGDRbM16krLMAzjrehM5g70Rk0Vpwiav2r3wKUlG3Dp8wfRuwYJmuL5/dS9WPjXmxizYQ2y9GzxZtwLj7LUrF69WnRfJfeTEqjNvXWr+4p2J2YYhnEWOtMFcWvSVoO7cOhqGyz4y2wh71zrR4QGFHbvZhh3wGMsNefOncPw4cOxatWq61rJl0RCQgIiIyMtS82aNR0+T4ZhGDXQSRfFrVFbFe7A5cwaSPhjAQymANxSYyOebDnX1VNiGM+11OzevRuJiYlo06aNZZvRaMSWLVuwcOFCYZHR6WxNtBMmTMCoUaNsLDUsbBiG8Sz3U3VXT0W4md7YvBjpeTGoH/0vRnYYB62GklFcj8Y/CLqwKAq2cPVUmPIiSTBmpkLKz4XPiJoePXrgwIEDNtuefvppNG7cGOPGjbtO0BCBgYFiYRiG8TQygsYgO+AJmDShDj1OWPTnIkAYq4AVM7eid43rM5ve2TYH59MbICb4CiZ1eRlBfjlwPRqEtb8P4c27Quvnz6LGk5EkmAz5yDiwGZk7vyvI3vZyUUNpXM2aNbPZFhoaitjY2Ou2MwzDeDqSJhQG3Q3OyXhasgEJk7cWGyCclheDxKzqCNRl49UuLyM2JBHuAAma6HZ3olJMFAL8NGBJ47mQhNEbJPgF3inWM3d+6/2ihmEYhlHPOjN188ci42n3kkEFGU/F7xsTfBUzez2G06mNUD/mENwBTUCwsNCQoAkP5FRybyCQXsaYKBiad0XWnvXldkV5tKj5/fffXT0FhmEY9ZFyEZkzBUZtNWQGvkyBI6oKmtad5uOvVX0xdHJT9K7xYLH7ZerDERaQYX5OQDqaVVaWdeoMdKGRwuVEFhrGewjw04jXlWKkDCmXvTv7iWEYxpeChEPyP0NY3kLVf3uShabaqr7YnTGoxHo0Z1JvwKD/bcAPRx+HW0LxMxp2OXkbGqvXtrywqGEYhnEzdFJB5pO2utMDYFNzY/DmlveQqY/CtrN3iEBhhvEUWNQwDMO4aeE9cj85E70xQNSiScyqgSphZzC+8zDu6cR4FCxqGIZh3AydqaDwnsZ5hfckCVi48w0cvtYGof7peK3rS4gITHXa8ZnSOXv+AqLrN8eBQ4ddPRW3hkUNwzCMm6G1dj+pGCAs93iieBrKeLLmy39fxOYz90GrMWDsrSNQI+KUasdmGGfBooZhGMbL3U+WjKfaPUTGU1FBczSpOVYdGC7uv9D2TbSqsh2+xPFTp7Hx9z9w4tQZhx/r2582oOOdD6Dqje1Qr+2t6NP/OWRlZ4vHPl7zFdr3ug9VmrTFzT3vxbJPV1ue17LrHeK2y70PC4vNPY8/LdZNJhPeWbAYTTv1QHyTNuh8z0P4ZfNWy/P0+nyMeX06Gt9ymxi3eedemL14meXxRR+uFPOp3uxmNO10O16Z/CYys8zz8UQ8OqWbYRjGG1HT/WRdYG/PUqpHs/66fW6IOYABrWYiNacS7rxhDXyFlNQ0PDdiAn774w/Ltu6dO+PDeQmIioxU/XiXE6/iuRHjMHXcSNzTqwcysrKwfdce4fr74tvvkTB3Ed55fSJa3NgY+w8dxvCJryM0OBj9+t6PX9d9jh4P9MM3H3+Axg0bIMDfnOa/ZMWnWLjsY8x5czJaNG2MT79ch8cHDcX2n75B/bq18f7KVfjpl9/x0YJZqFGtKi5cuowLFwvTpbVaLd6ePAG1a1bH6bPnMXrKm5jy9my8O+1VeCIsahiGYdyMpLBvhbCpqPvJ4m4qKLBXEpRg9WCTj+BrkKDZ/Oe/AD4lGwiALdj85zA8O3wCvlrxnurHu5J4FQaDAff0vh21qputcE0bNRS3b819D29MGI17e98u1mvXrIEjx05g+edfClFTKSZabI+JjkJ8XCXLmAuXrcTwQc+g773marxTx43C1h27sHjFJ5g19VWcv3gJ9evURod2baDRaCzHlXnp6f6W+7VqVMekUUMx6rU3WNQwDMMw6iBpwmDQmb/sHEVOfghWHRiGx5svQIh/FnwNcjmZLTQkaJ4o2PoEjEYJv/3RX7iiyNKhJs2aNELXju1x610Ponvnjrjt1o64/86e8Pf3x6mz5zBswhSMmPS6ZX+DwYiI8LASx0vPyMSlK4m4pW0rm+3t27bCwf+OivuP970fDwx4ATfdfi96dOmE3t27imPL/L5tO+Ys/hDHTp5CRmamOGZuXh6yc3IQEhwMT4NFDcMwjI9hNGnx7vaZ2HWhO86m3oBp3Z+Fr3HqzLmCe2Shsaar+HvyzFnVRQ01Xl738QfYuXsvNm39E0s//gxvvrsAn3+wQDw+d8YUtGvZoshzKhb62rLZjdj7+88izmbzth14euhodOt0C1Yumi0yqh57bgieeeIRvPrKUERHRWLH3/9g6PjJyM/PB1jUMAzDMBXB37AXIfrPofdrhZyAfg45xsf7XhGCxl+bhydazIcvUrd2zYJ7W6wsNcRm8bde7VoOOS65gG5p11osY4e+iBZdemHn7n9QNb4yzpw9j0fuv6fY55E1hzAaC+sGkRWHnrdj9150an+TZTuJpjYtm9ns9+A9d4jlvjt74qGnXxTxRHsPHoJJMuHNiWNEbA3xzQ/Xx1x5EixqGIZh3Ah/40HRIkErJVZI1FDGEwUIYxWwYiZ14DZv33D8YXxz+Blxf9gtE9Go0j74Ig3q1hFBwRRDQy4ns4VmM3S64ejasbPqVhri7737sfnPneh+a0dUqhSD3Xv341pyChrWr4fxw1/G+GlvISI8HD26dkKeXo+9B/5Falo6Bj87AHGxMQgOCsIvW7ahWtV4BAYGIjI8HEOfH4iEue+hbq2aaH5jI6xa+w0O/HcYS2e/Zcluio+LE0HEJFwo+4piciIjwlG3di3k5xuwdOVnuKNHVyGOln/+BTwZFjUMwzBuhFYqyHzSVlU942n/5fZY8vdr4n6/ZgvQpfaP8GUoy4mCgimGRoYEDW13BOFhYdj+124sWf6piF+pWb2aCA7u2a2zeDw4OAgLPliByW+/K+JZbmx0A14aaJ6bn58f3po8Hu8sWCKypDrc1Abff7YcgwY8IWJrXkuYiatJyWjUoD4+e3+BRZSFhYZi/gfLcfL0GWi1OrRp0RRffPieEDjNmzTC9EljMG/pR5g2ax463tQWr40egZdGT4SnopEkSibzDdLT0xEZGYllWw4hJCzc1dNhGIa5jsjskQjJX4v0wPHIChpst3WGGlbK9WisG1ZeSK+DMRtWIys/El1qf49RHcY4u62UavjFVEX8o5NQs0ocAioYc0JQUDDF0JDLyREWGkYZeqMJ5y5fxZU102FINheglMnON6Lf2mNIS0tDREREiWOwpYZhGMYLCu/JBfYWL9mAoZO3XteBO88YhEC/XNSIPImh7Sd5rKBxBCRkWMx4ByxqGIZh3Aidxf1kn6gRFppVfXEpA9cJGqJe9GHM6vUIdBojAnR61ebLMO4Et0lgGIZxFyQTdCaz2d2kQosECi64lCFn+QCxIYmICk6q8LgM466wqGEYhnETtFIyNNBDggZGTZUKj/fN4acx7Kfv8McZc7VZhvF22P3EMAzjJpi0lXA54jC0piuAxlyXpLzsPN8dK/eOhgQtUnNjVZsjw7gzbKlhGIZxIyRNKIy6enYFCMs9nuQO3CeTm+DdP2cKQXNng89xT0NqBcAw3g+LGoZhGA/FkvEU84UlhTspOw5vbnkPecYQtKqyDc+1ncGZTozPwO4nhmEYNyFY/zkCDHuR43839P5FexKV1IF7g7DOUMXgPEMQZvyxCEk5VVAj4gTGdBoJP63BafNnGFfDlhqGYRg3IdCwRbRI8DOZOyyXKWgyBglBI/PTscdwPLk5wgNS8GqXlxAWkOGEWTOM+8CihmEYxk3QmS5WKJ37vsYr0bfJUkzoPBRVw+Uu1Iwvc/b8BUTXb44Dhw675Xhqw+4nhmEYNxM1Rk35RI1WI+GpVnNUnhXjyVSvWgWHd2xCbHQUfAG21DAMw7gDUj600hW7qwkfutoGC3a+iXxjxVLAGc8kPz+/1Md1Op3oyk0NMd0Fvb70OVcEFjUMwzBugE66Ag0kSPCHSVOp1Iwn6sBN8TTzpu3DjC0L8cvJvvjy0CCnztebOH7qNDb+/odobOlIVnz+JZp06A6TyWSz/fFBQzFknLl7+o8bf0PX+x5BlSZt0arbHXh7/mIYDIXB3uT6+XDVGvR7YSiqN7sZ7773AVLT0vD8yHFocFMXVL2xHdp2vxur1q4r0V3039HjePS5wajV8hbUbNEedz46AKfOmN2VNLd3FixG0049EN+kDTrf8xB+2by11PPatnMXejzQT+zf+Jbb8Po7c2zmfM/jT2PM69Mx4Y23Ub9dZ/Qd6Lj3qvtIN4ZhGB9GK7uetFUBjbbMjKdjKwdg89HPkKGPRv2Yg3iwyYdOnrHnk5KahkEjxmHjH9ss23p27oSl895GVGSk6sfrc1cvjJuWgD+2/4WunW6xzOHXLdvwxbL38Oeu3Xhx9CS8PXk8OtzUBqfOnsOISdPEfuOGvWQZ5+1572HK2BFIeHUsdH5+mD5nIY4cP4kvP1yM2Jgo0XE8Jzev2DlcvHwFd/cbiFvb34RvP/kQ4eGh2Ln7HxiMZhGyZMWnWLjsY8x5czJaNG2MT79cJ0TX9p++KbbpJ433yLOD0a/vfVg8azqOnTiF4ZOmIigwEOOHv2zZb/XX3+Hpxx/Fz198DEfCooZhGMYN0FEV4RLiaeR6NI+u6ivq0dxe7WHM3PY+zqc3QGzwZUzq/DKC/HJcMGvPhgTN7j93gEoTUgL9FgDD/tyBF4aPwxcrlqh+PBJKt3e5FWv/96NF1Hz70wbERkejc4eb8eCAFzDixWfRr+/94rE6tWpi4sjBeP3tOTai5qH77sITDz1gWT9/8TJa3NgYrVs0Feu1alQvcQ7LPl2NiPAwfDjvHfj7m12WDerWsTy+cNlKDB/0DPrea26tMXXcKGzdsQuLV3yCWVNfvW68Dz9dg+pV4zHzder8rkHD+vVwKfEqpr4zB2OHvgit1izQ69WpjWnjR8HRsKhhGIZxA3ID7sVl/+7QSJnXWWfI3XRpyQYkTN6KXtVn4/2/J2Pv5U4I1GVjUpfBiA256rJ5e7TL6Y9tQtA8UbCNbiWjEf3/2CZcUcVZJirKw/ffjeETpwqBEBgYgC+/+wEP3nOH+PI/ePgodu7ei9nvLbXsbzSakJuXh+ycHIQEB4ttrZqbxYvMM48/ggGDR2Hfv//htls74u6e3dG+batij09uqA7t2loEjTXpGZm4dCURtxR5Lo118L/iywwcPXESN7VuKQRN4f6tkZmVjQuXr6BmtarmOTe9Ec6ARQ3DMIwbtUigpSjVVvXFpQyIisHfH30CPx3vBw1MGNVxLOrHHHLJXD0dOYakaInDrgW35MJxhKi5o0c3SNLr2PD7FrRu3gzbd+3BjEljxWNZWdnCZXNv79uvex65c2RCC8SNTM9unbF/y3oRF7Rp23b06f8cnnvyMbwxcfR14wQHBcEVhITYztlRcKAwwzCMB1Er8hjCAlLxVKt3cUuNX109HY+lbu2a4pZcTtZsLritV7uWQ45L4uTe3j3w5bc/4Kv//Ygb6tVBy2ZmK0aLpk2EBalenVrXLbIbpyQqxcYIt9XS2W9hxqtjsXLN2mL3a9q4Ibb/vbvYrClyS1WNr4wdu/fabCfrUaMbiu9HRu6mXf/sgyRJVvv/g/CwUFSvEg9nw5YahmEYNyAiZxI0kgGZgS/CqKtb4n4t4v/CgrvuQ3QQu5wqAsWRUFAwxdCQy6lrgaAZrtOhZ8dbHGKlsXZBPfbcEBw+dhyP3H+PZTvFoDz2/BDUqFYV993R0+yS+u8I/jt6DK++MqzE8WbMWYhWzW5E4xsaIE+vx/pNW4TYKI7n+/fD0o8/w7PDx2Lki88JIbNr7360bdkMN9Sri6HPD0TC3PdQt1ZNNL+xEVat/QYH/jssxFJxPPvkoyK4eOzUGWLs4ydP46157+HlZ54qU4g5AhY1DMMwbkBw/v+glVKQFTjwuscSU8Jw+nIAZAdETDALGjWgLCcKCqYYGhkSNLTdkXTp0B7RUZE4dvK0CPqV6dGlE1Z/sBDvLFiCee9/JGrLNKxfF/0fKWyFURwB/v6YNmsezp6/iKCgQHRo10YEAhdHTHQUvv10Gaa8NVukWut0WjRr0sgSRzNowBMitua1hJm4mpSMRg3q47P3F5Qo8qpViccXHy7C5Ldmo/OahxAdGYknH34Aowe/AFegkaxtRl5Oeno6IiMjsWzLIYSEhbt6OgzjkVw6cxJXzp9BfM06qFqrZIsCYwdSDqqmNxR3L0cchKSJtGQ8xSx/FPUfHo2MnBBM6DwEzeN3wdfxi6mK+EcnoWaVOAToKm4NoKBgiqEhl5MjLTRM6eiNJpy7fBVX1kyHIfmSzWPZ+Ub0W3sMaWlpiIiIKHEMttQwDKOIzLQULJw4Avu3/2bZ1qJDdwxJmIewCN8owe7wnk8Ig4QIS8bTxcUb0P1JP1xOr4JQ/zREB19z9VS9EhIyLGa8Aw4UZhhGESRoDu6kAEJKgj0rbml94YThrp6ax6MzXRC3kn8o3j1wlyiwRync81/dgGNXb4BWY8C4W0egRsQpV0+VYdwattQwDKPI5WS20NhW9TCZJOzf3h+Xzp5iV1QF0ElmS82x9Cz8Vfsx0QLhy7eXYtUBc3PKF9tNQ8sqO1w8S4ZxfzzGUrN48WK0aNFC+NJo6dChA3766SdXT4thfAKKoUEpVT2unDvt9Dl5EzqTOX4g1WAuiHbwSBDm7UgQ9+9rtAK9G3zp0vkxjKfgMaKmRo0aeOutt7B79278/fff6N69O+6//378+++/rp4aw3g98TXkeIPiq3pQ0DBTfrRSkrhNLei0/dm3sdAbg9Cu2u8Y2Gqmi2fHMJ6Dx7if7r33Xpv16dOnC+vNjh070LSpbclohmHUpWrteiIo+ODOYcLlZLbQbIZWOxzN2ndn11MFSQ9+ExlBE/HH2T5AFDB15AVUSv0Y9zdeDp3WtqMzwzBeYKmxxmg0YvXq1cjKyhJuKIZhHA9lOTVrT7Us+lNdW3FL67SdqTgz9/fFmgkZIp5m1dyt6Nd8EUL8s109LYbxKDzGUkMcOHBAiJjc3FyEhYVh3bp1uPHGkptk5eXlicW6Tg3DMOWD0rbHL1opgoIphobr1KgD1aOZuvljdP1xPsbe1AIPr+qC3jXYOsMwXm+padSoEfbu3YudO3fipZdewoABA3DoUMnN3BISEkSxPXmpWdPc64NhmPJDQqZVp9tY0KgkaFr0nIvXg3XQ1FyN1d9F49DVtq6eFsN4LB4lagICAtCgQQO0bdtWCJaWLVti3rySTd8TJkwQ1Qfl5dw5c1dWhmEYV0MF9qhi8C8vf4CwyvvQqdM3eKzZIq4Y7MNQz6TO9zxU4XG27tiF6PrNkWaHd+LlMZPwxKCS+0t5Ch7lfiqKyWSycS8VJTAwUCwMwzDuZqEhMhJewmcr9HitNZCTFoFHm77n6qkxLmTIcwPxwlOPV3icm9u0wuEdmxARrrwdUMLk8YAXdE3yGFFDVpc777wTtWrVQkZGBj777DP8/vvvWL9+vaunxjAMYzeGzEjcl/A8mt30hVivJCUiT1PN1dNiXEhYaAhASwno9fkICDCn/ZdGQIA/4uMq2XXsSDsEkDvjMe6nxMREPPXUUyKupkePHti1a5cQND179nT11BiGYeyCfhCfXPQujl+IQ93aBYULs3SunpbPcvzUaWz8/Q/R2NKRrPj8SzTp0F14Gax5fNBQDBn32nXuJ9klNGvRUvG8m3qaS5vs3L1X7FelSVvcdv+j+GHDr8LddODQ4WLdT5+t/Qa1W3XEr1u2oX2v+1Cj+c14aOCLuJx4tUT3E82ROoW3ue0uxDdpg2a39hTzkJny9my063EPqjW9Ca263YHpsxcgPz8frsZjLDUffvihq6fAMAyjChoNENd9NUKvNEbPbkfFNinLYz6OvYaU1DQMGjUOGzdvs2zr2bUTls55G1GRkaofr89dvTBuWgL+2P4Xuna6xTIHEhtfLHsP2//ec91ztmzfifCwMHy90iwo0jMy0e+FIejZrTM+mPM2zl24iIlvvlPmsXNyc7Bw2QosmZUArVaDQa9MwGsJs8QYxTF15lx8vOYrzJg0Fre0a4PLV6/i2InC3mPhoaFY9M6bqBofh3+PHMOIia8jLDQUwwc9A1fiMZYahmEYb4mnoQDhYw9JuLfxV/DP2Si2mzLLdisw6kKCZuc/OzBxIrBmDcQtrb8wcpxDjkdC6fYut2Lt/360bPv2pw2IjY5G5w43F/uckOBgzE+YiiYNG4iFnqvRaDBvxutofEN9IW6GPj+wzGPn5xsw+43JaN2iKVo2uxHP9e+HLX/uLHbfjMwsvL9iFaaOG4V+fe9H3do10aFdGzz1aF/LPqOHDEL7tq1Qq0Z13Nmjm4gH+uZH14eD8E8DhmEYJzHl86n49ON9uPTBBuxZ+iDuqb0eGn+zK0JiUeN8l9PmbULIyFEM5lsjZszYJlxR9evK7UHU4+H778bwiVMxa+qrCAwMwJff/YAH77kDWm3xNoYbG91gE0dz/OQpNG3UEEFWSTBtWjYr87ghwcFCnMhUiYvD1aTkYvc9euIk8vR6dO3YvsTxvv7+Z7y/chVOnz2HrOxsGAxGYVFyNWypYRiGcYJ15o3/jcSFBQtxW/Lv2LN0uOWxrK/rIv2jhjBeDXLpHH2NU2fMJT5atrTdLq+fPHPWIce9o0c3SJKEDb9vwfmLl7F91x48fN/dpYoRNfDz87vOBUrzKI6goNKzhv/asxcvjBovrESrP1iEzd99iVdefh56N4ipYVHDMAzjYEEz6ZufkL7gA2TnBqJh7B5UDr1ou5NBC0gaV03RJ5GtFvv22W6X1+vVplYg6kMWlnt798CX3/6Ar/73I26oV0e4g5TSoF5dHDp6DHl5esu2f/ar29i5fp3aCA4KwuYS3FN/7dmHmtWrYvTgF4Q7iyxa5y6aO827GhY1DMMwDiQ32x8n5i3AhWtRqFszD2M6jYJOa3T1tHyeBnXriKDghQt12LiRMmwhbmmdtjvC9WTtgtqwaQtWrV1XqpWmOB669y6RmTRi0lQcOX5SBBlTALDF/KKS8KKAX8pwWv31d8Kqteufffjki6/F4/Xq1BJWpq/+95N4jOJvvt/wK9wBjqlhGIZxEJS5O+eVu5B9uiFiIzIx+7XL0G7KsDyuq56FwBbJMFwIgX5/rEvn6otQlhMFBVMMjUzPrreI7Y6kS4f2iI6KxLGTp/HQfXfZ9dyI8DB8vnQhXpn8Brrc+xBubHgDxgx5Ec+PHIegwADV5jhmyCDodDrMmLsIlxMTER8Xh6cff0Q8dtftt+Glp/tj7NQZ0Ov16Nmti9j/rXmL4Wo0UklONS+EGlpSD6hlWw4hJMw7Cg0xDOO+rF4Qge+WR0Djp8fvsxcjuE4vXPr8QcvjAc2TENThKvKPRyDnNy68pxS/mKqIf3QSalaJQ4Cu4g4HCgqmGBpyOTnSQuMovvj2e1Hn5sze7cJt5KnojSacu3wVV9ZMhyHZ1p2VnW9Ev7XHRMujiIiIEsdgSw3DMIwD0OcC+/40f8HMDhuL4DqDbAQNoQkziFtTJn8UuxISMp4kZsglVLtWDVSLr4yD/x3B1HfmoM9dvT1a0KgF/ycxjA9w6cxJXDl/BvE163B3bScRU/VzRI/8Em+eG4LtW5qiXhFBQ2jDzNkipixO52aUc+XqNeEWSrx6DfGV43D/nb3w6iue34xSDVjUMIwXk5mWgoUTR2D/9t8s21p06I4hCfMQFhHl0rl5K/o8IKaKucDe0iUbsGLVVvSuMbvYfbWhZkuNxJYaxg4oiNfVlXvdFc5+YhgvhgTNwZ17AXwKgOpufCrWF04orJPCqEdWhgYv9s3FthZGXFy8QbibShI0hEa21HDhPYZRBRY1DOPFLiey0JhM8wE8AYDqcjwBk2me2H7pbGEfF6biGPKBRZPykXuxAT4NfBHpmWV8vGpN0IaYU7vZUmMnlN8iSfCZLBcfQbJ6bcsLixqG8VIohsZMlyKPdDU/fq6gOzRTYegzeOXMKOzdWgfagGx8O+MDRIbbdmIuiibYCClXC8mggZTHHbrtwZiVBpMhH3oDyxpvQm+QxOtqzEwt9xj884BhvJT4GnI2x5YCS43MZvPjNeu4ZF7eyPrVYfh1bRg0Ggl1XxyP1jdosbuwHE2xSFn+yPi4IaAj8cPVhO1B0ucg48Bm+AXeCcREIcBPw1fQg5EKBM215FTxukr5ueUei0UNw3gpVWvXE0HBB3cOg8kkFVhoNkOrHY5m7btzFpRK/LM1CJ+8GynuDxy/GfsbbQLQAytmUoCwggGMbDAvD5k7vxO3huZdofXzV62aLuMCJLOFhgSN/LqWFy6+xzBeTGZ6qggK5uwnx5B6TYtXHqiCnCwtej6yHx/+cTcuv7+hQNCUHCDMqIfGPwi6sCgWNZ6MJAmXU2kWGqXF91jUMIwPQEHBFEPDdWrU57evQ7Fm3X9InPMVDuQ+d12BvZIIaH0NflVyoD8UBcMZ/jximNLgisIMw1ggIcNixjEduP9p8DFSFvTAsClN0buGMkFD6OJz4FczC/knWdAwjFqwM5dhGMYOyLb93fJwmLAOUzd/jL9q98CezEF2u5u0lhYJXKOGYdSCRQ3DuKiGzN5tm7hWjAeyblk4Vi+IxPiHH4fJUH5jt9wiQcpigznDqAX/NzGME+G2BZ7N9vXBWLvYnOl0/7N/Y4ef2dpiN/5GaALMdWzYUsMw6sGWGoZxIty2wHM5fiAAS16PEffvejIDvfvtL/dYsuuJiu/BwB/DDKMWbKlhGCe3LTALGrkYHrUtkLB/e3/hiuJgXvfk2iUd3h0Vi/w8DVp3zsHjw9Ms8TS7MwYpzniS4e7cDOMY+CcCwzgJblvgmeRkaTBrRCzSknSodYMe4957H13v7yAEzdDJTe0WNAI/E0y5Wpi45xPDqAr/RzGMk+C2BZ5JeooWuTlaRMYaEfni3ejU+zAuLdmAhMnlL7BnOBWBzFMRVClM9fkyjC/DlhqGcXLbAq12WIEL6py4pbYFtJ1dT+5JfA0jpq1MRLXhj2Bvm6ZC0JB1RpWKwRJXwWUYNWFRwzBOhLKcmrVvBaA/gFriltZpO+NepCUVfjxWq7cKIbWOunQ+DMOUDbufGMaJUNr2+EUruW2Bm3NgZyDeHRmLAWNScdsD2aqPH3LHOdGYO3dbPEzpAaqPzzC+CosahnEB3LbAfblwyg/zxsRCn6vFob8D0a2P2qJGgq5aNjR+kqhOzDCMerD7iWEYpoCMVC1mDa+E7EwtGrbMw/OTUyzNn7+ZmCpuqQN3RdAEGS2CRuKUboZRFRY1DMOojie2gTDkA3NGx+LKeT/EVTNg5LtJCAgE3t1/J1p3mi8ChCmFu6IBwhq5PUK2H2DiQGGGURN2PzEMA19vA0FWkw+nR+PwnkAEh5owZt41VK+/qrDA3hJzgb3eNSp+rMJGlvzxyzBqw5YahmHg620g/vkjCJu/C4VGK2HYW0lo3O4TYZ2pUIG9MhtZsuuJYdSGfyowDANfbwPRunMuHhmcJqw0ne5ZYbbQrOorWiD0rqGeoCE0oWypYRhHwf9VDMM4rQ2Eu4oaCgbu82yG045HLRIk7s7NMKrD7ieGYRzQBgJu3wYiOVGLD96IQm62c4N183ZWRubHDaE/GO3U4zKML8CWGoZhVG0DcXDnMOFyMltoNos2EM3au1cbiNwcalJZCacPByA3W4uhCckumAVnPjGM2rClhmEYn2oDYTIBi1+LFoImPMooYmlkwqI/L8x4yjBnPDEM4zmwpYZhGJ9qA/HFogjs+i0Efv6SqEVDDStlQSMynlb1LahHo76g0YTmI+SuczClBSBngwr54QzDeKalJiEhATfddBPCw8NRuXJl9OnTB0eOHHH1tBiGKQYSMq063eZ2gmbL/0Lw3fIIcf/511LQuLVe3Fe7wF5p6dy6aD10MXkOGZ9hfB2PETWbN2/G4MGDsWPHDmzcuBH5+fno1asXsrKyXD01hmE8gMN7AvDBG+bg3D7PpqPzPdnCOkOChtxNJGjMBfYcI2gIjaXwHmc+MYxPu59+/vlnm/UVK1YIi83u3bvRpUvRFFKGYRhb/AMlhEeZ0Kh1Hh56Kd0lc5AL73GNGoZxDB77n5WWZg7ui4mJKXGfvLw8scikp7vmg4xhHMn+Pzfj+MF/0KBFG7S4hQV+SdRvmo83PklEWIQJWhfZqLmaMMM4Fo8UNSaTCSNGjECnTp3QrFmzUuNwpk6d6tS5MYyzoEDc1556EJlpVy3bwiLj8Man6xBfXa4Z49tQk8rL5/xQo57Z7RMbbw4KdhVcTZhhHIvHxNRYQ7E1Bw8exOrVq0vdb8KECcKiIy/nzp1z2hwZxtGYBY3eps8Srb/25AOunprbNKlcOTMKrz5ZGbt+Cypxv28mporbFTO3Os9SwzE1DOMQPO7nwpAhQ/D9999jy5YtqFGj9JTIwMBAsTCMN7qczBYa2z5LgITMtP7Yv2OLz7ui1q8Ow69rw6DRSNDqrn+cAoRJ0FCA8J6lFCC83uFzkvJ0okUCW2oYxsctNZIkCUGzbt06/Pbbb6hb171SRRnGmVAMTWl9lo7v3wNf5p+tQfjk3Uhxv9/wNLTtmmt5rLiMJ2eR/UMt0SLBlFKy5YhhGB8QNeRy+vTTT/HZZ5+JWjWXL18WS05OjqunxjBOp0Gz1qX2WaKgYV/l7DE/LJwQA8mkQbc+Wbi7f6blMUuBvdo9RD0arhjMMN6Fx9hAFy9eLG67detms3358uUYOHCgi2bFMK6hRceuIig4M22wcDnJfZaAIWK7r7qe0pK0oqdTTpYWN7bLxTMTUkQHbpsWCKv6ihYIjqgYzDCMa/HzJPcTwzCFUJYTBQVTDE3R7Cdf5afPwnDtkh+q1MrHiJlJ8HOjeFy/BmkIbJ0Ew6lw5P0d5+rpMIxX4jGihmEYWyhte+mmPSIomGJouE4N8MjL5lpUXe/LRlike/0Q0kXqRYsE42VzWjfDMOrDooZhPBwSMr4uZmQoy+mxoe5ZZLOwRo0bmY8YxsvwmEBhhmGY4vhzfTCWTosWhfZKwhJP48IAYa5RwzCOhy01DKMiv69bjX93b0fTmzuh232PwBu5dOYkrpw/g/iadVzehfvY/gC8PyUG+XoN6jbWo+cj1ze4tWQ8rerr0A7ciptZZvHHrqdxIV2Py5l6VA0PQLXwALcZi7ke/u9iGBU4eWg/pgzsC6PBXA9l249f48M3J2HaJ1+jbqPm8AYy01KwcOII7N/+m2Vbiw7dMSRhHsIiopw+n6sXdXh3VKwQNG265KBH3yy3KLBXPBK0oXIzS7bUeAoZeUbM2XERuy8WvrfaVgvFqA7VEBagc9lYTMmw+4lhVMAsaAJtWhbQ+uT+3pM2TILm4M69NudI6wsnDHf6XHKyNJg1IhbpyTrUbqjHkBnJNlWDXVlgrzg0gSZo/M2ByxJbajwGEiFH07MwcSKwZg3ELa3P3n7RpWMxJcOihmFUcDmZLTSLCloV1Cy4XSi2//7dF/AGlxNZaEym+TbnaDLNE9svnT3ltLmYjBDF9c4dD0BUJSNGz01CUIh7ZTpdR4ARxmuBMKYGAEb+2PUEyE1EVpWhw4CePYHKlc23tE7bL2boXTIWUzr838UwFYRiaEprWfDvX9vg6VAMTWnnSB3DncXn8yPxz9Zg+AdKeGXONcRWcW3nbSVIGQHI+rousr6o5+qpMAqhuBeiZUvb7fL6JTuEiJpjMaXDooZhKkjTth1KbVlAQcOeTnyN2qWeIwUNO4vmt+QiJNyEl6Ylo37TUlKeGKYCVAkzB/Hu22e7XV6nQF9XjMWUDosahqkg3R54DDo/alA4uCDe5FzB7RCx3RuyoKrWrieCgrXaYTbnqNUOF9udmQXVokMe5n53Cbf0LL3vGwUIUzuEFTO3Om1ujPdQPSJABPIumA9s3AgkJppvaZ2225O5pOZYTOloJB/qP5Ceno7IyEgs23IIIWHhrp4O40WcOnJABAXL2U8ECRqvyn5KTxVBwa7Ifrp42g8aLVC1VtnVeOUAYRI0rg4QlgnqfAm6KjnI+7sSDKciXD0dRiGZeqMI5FUjY0nNsXyR7Hwj+q09hrS0NERElPw/xKKGYVSEgoIphsar69ScPSViaJxVpyY9RYspAyojM12LcQuuoUFzvaICe66qR1McIfeegV/VHGT/Ug2GkyxqPA0K5KW4FzVqy6g5li+RrVDUcG4hw6gICRlvFTMyJGSc5W7K1wNzR8fiynk/xFU3oHKN4i01coG9xUs2YOjkrW4laGyqCXM6t0dSTUUBouZYzPVwTA3DMG4J2ZA/nB6Nw/8EIjjMhDFzryEi2lSihabaqr5i3d0EDTQS931iGCfBPxsYxk1bCCgdy53aFqjJ/1aGY8v/QqHRShj2VhJq1PfM7taaYIOIB5JMgJTNH7kM40j4P4xh3KyFgNKx3K1tgZr89WsQVs+PFPcHjElFy4558FS0BT2fhOtJ0rh6Ogzj1bD7iWHcrIWA0rHcqW2B2m6nX78KE/d7PZqJXo9e39PJk9AUxNOYstj1xDCOhi01DKNSCwGzuKAWAihoISBh//b+IltIqVtI6VhqHtPd0GggKgVv/CIMd/TLLHXf6ztwu0cKtw1GjWiRYEqm3mAMwzgSttQwjBu1EFA6lju1LVCzp5NMQCBwd/9M6PzKFjTUsNLdUritMZwJFy0ScrdWcfVUGMbrsUvU/Pfff5gyZQq6d++O+vXro2rVqmjRogUGDBiAzz77DHl5nuv3Zhh3aCGgdCx3alugBiYTMG9cDD6fHyHuKymwd/fpC5YO3O4qaBiGcUNRs2fPHtx+++1o3bo1tm7divbt22PEiBF444038OSTT4Lq902aNAnVqlXD22+/zeKG8SnUbCGgdCx3alugBl8sisCu30Lw06pwXDjpV6p1xh0rBjMM40ExNX379sWYMWOwdu1aREWVnFWxfft2zJs3D++++y4mTpyo5jwZxq2hjCNzC4H+lm3N2pszkRw1lprHdCWbvwvBd8vNFUJfmJyCmg08M3W7JMIePQHJpEH2zzVEt26GYVwsao4ePQp//7Ij9zt06CCW/HzunMv4FpRCPX7RSlVaCCgdS81juor/dgdg2ZvR4n6f59Jx693Z8Cq0JmgjCz4P9RzCyDBuIWqUCJqK7M8w3oKaLQSUjuXMtgVqcvmsDnNGx8Jo0KB9z2w89GI6vA25krBk0EDK46aFDOOWKd27du3Cpk2bkJiYCFORqL7Zszlgj2GY0jHkA++OrITMNB3qN9PjpanJ0CowZFA7hG8mpmL3kkFYMZN6PMGtkXs+mdsjcOE9hnE7UTNjxgy8+uqraNSoEeLj46GhohIFWN9nGIYpCT9/4IEX0vHVkgi8MvsaAoLKfo4lQHiJOUDY3QWNTTXhTC4JxjDOwO7/NAoE/uijjzBw4EDHzIjxeZzdy8jXeyy5io69c3Bz9xwhcEpDrkfzqDsX2CsBriZcyIV0PS5n6lGVu1Qz7iRqtFotOnXq5JjZMD6Ns3sZcY8l5/Pn+mA0aZOH6Diz27osQUPWGXI3UT2ahMnkbvIs97ZWjqnxYUtNRp4Rc3ZcxO6Lhe0u2lYLxagO1RAWwHFGjLrYHY4/cuRILFq0SOVpMIzzexn5eo8lZ/PPH0FYNCkGr/WvjLSksj96yEJDVFvVV9x6mqAhTNl+okWCMdV3LRMkaI6mZ4GqfKxZA3FL67O3X3T11BgvxO6fD6NHj8bdd98tKgrfeOON12U6ff3112rOj/ERnN3LiHssOZezx/ywYEKMqNfSslMuImIUlA32AvR7KonFl11OZKEhIdOzp3mbfDtjRhYuZujZFcW41lIzbNgwkfnUsGFDxMbGIjIy0mZhmPLg7F5GvtxjydmQVWbW8ErIzdbixna5eHp8qmhayXg/FENDtGxpu11ev5RhfpxhXGapWblyJb766ithrWEYtbDtZSRbRBzXy0jp8Zw9L29Dnwu8OyoW1y77oUqtfIyYmVRmHI33IBXc+q6CqxJmtsLs21dooZHXCQoaZhiXipqYmBjhemIYNZF7GR3cOUy4dsyWkM2ilxGV/lfbxaP0eM6elzchScD7U2Nw/EAgQiNMGDMvCWGR8he9soynvzww40lGG61HaJ/TMCYFIfs7WRz7FtUjAkRQ8IL5WRYLDQmaBfPNwcLsemJc7n56/fXXRafu7GwvK2fOuBzKJmrWvhUA6mVUS9zSuqN6GSk9nrPn5S1kpWtw/oQ/dH6SsNBUra2spxNlPJGgoYwns6DxvABhQhOaD42/BI2fb8QPlQRlOTWMCMWMGcCjj1IsDcQ6bWcYtdFI1GLbDqhT94kTJ0Rn7jp16lwXKEwdvd2V9PR0EfezbMshhISFu3o6TAk4u5eR0uN5co8lV5GdqcGxfQFo2SlPkXWGKgZ7Swdu/8apCO5yGflnQpGzviZ8HQoKphgarlPDlIfsfCP6rT2GtLQ0RESYG+Cq4n7q06dPuSbEMEpxdi8jb++x5GxysjQIDjX/VgoJk3xS0Fi3SJBEiwSGhAyLGcbR2C1qyPXEMAxTHFcv6jBlYGXc3T8Ddz2Z6dNZTpZqwixqGMZ9Y2qomeXOnTuv207b/v77b7XmxTBux/4/N+PrpXOxf8eWCu1DUP2bvds2CZeWM3DG8cjVNGtEJaRe02HrDyHI9/FsXbmasCnLd6sJM4yzsfu/bfDgwRg7dizat29vs/3ChQt4++23ixU8arFlyxbMnDkTu3fvxqVLl7Bu3Tp2hzEOh+JoXnvqQWSmXbVsC4uMwxufrkN89dqK93HnVhAVxWSkyssxOHfcH1GVjBg9NwkBgfBpZEsNu58Yxo0tNYcOHUKbNm2KDSCmxxxJVlYWWrZsyW0aGKdiFit6mzYJtP7akw/YtY87t4KoKKvmRGLv1mAEBJkweu41xFYx2vV8iqehHk/eEk9DmK4FiRYJpgwWNQzjtpaawMBAXLlyBfXq1bPZTpYTPz/HmlnvvPNOsTCMsyB3ktn6YtsmgQqrZab1N7uZTFKZ+7S4pYvbtoKoKL+sDcVPn5mzCV+aloJ6N5otFEqwCRBe4j2Chsj5tbqrp8AwPofdlppevXphwoQJIq1KJjU1FRMnTkRP65KRbkBeXp5I47ZeGMYejh/8p9Q2Ccf371G0jzu3gqgIF075YcXbZjfWI4PT0P72HMXPtRTYq91D1KPxJkHDMIyHiJpZs2bh3LlzqF27Nm677Tax1K1bF5cvX8a7774LdyIhIcGmL1XNmlwrgrGPBs1aF9wrGvhrbpPQoEUbRftc33IBTm4F4ZjjVa9rQP9XUtH1vizc/0yG4udZF9j7ful6jy2wVyIaSmm3qwQYwzAqYLe/qHr16ti/fz9WrVqFffv2ITg4GE8//TT69et3XSE+V0MWpVGjRlnWyVLDwoaxhxYdu4qA38y0wQVfUuY2CcAQsZ3cSoSSfdy1FURF6f1YFiQpS3H6NlloiGqr+iJh8lb0rgGvI6BpCgJvugr94SjkbY939XQYxmcoVxBMaGgoXnjhBbg7FP9DC8NUBMpgooBfio8pmtlkzz4EZR1RkC7FtMiQwHBkKwi1j0ep2l++F4n7nklHWITZGuHL9WiKQxNqEC0S2FjDMG4oanbs2IFbbrlF0YDUE+rUqVNo2rRpRefGMG4BpWQv3bRHBPxSfAy5k2Triz37EJRGPX7RSqe1XFD7eNRU5cPp0djyv1D8tycQ01YmsqApBq4mzDBuLGr69+8vsp2ee+453HXXXcJSUxRK5/7000+xfPlyUa/GEaImMzMTx48ft6yTeNq7d6/oHF6rFjUaZBjHQSKlOKFi7z7u3AqiLP63IlwIGq1OwkMvprGgKauaMBfeYxinoug/jgTL4sWL8eqrr+Lxxx9Hw4YNUa1aNQQFBSElJQWHDx8WguOBBx7Ahg0b0Lx5c4dMlioWU2CyjBwvM2DAAKxYscIhx2QYxsxfvwZh9YJIcX/AmFS07Fh2T6cSM55W9fXoDtyKqwmzpYZh3LtLNwmLrVu34syZM8jJyUGlSpVE4T0SG2QxcWe4S7dnQPVVKB1ZDdcM1ZmhlOuS3EFqj6V07u56jiVx8pA/pj0XB32uFr0fy8CAsYUlHezJeKICe5TxZE/69oV0PS5nqtPdWelYFTqmRkL4c0eEFSvjkwaQcthawzBu26W7Xbt2YmEYdy7pr7RtgVpjKZ27u55jaSQnavHuyEpC0LTsmIsnR9knaKwL7A2t3xS9FQqajDwj5v55EX9fzrJsa1clFCM7VUNYgM6uOdBYc3ZcxO6LhWO1rRaKUR1sx1K6X5lBwhpAMmog5dg3T4ZhnFynhmEchZol/ZW2LVBrLKVzd9dzLI2cTC10fhJq1M/H0IQk6PzKKWjsdDeRoDlxJcvq7CDW52y7aPc5kFA5mp6FiROBNWsgbml99vaL5dqvVDQS8s+EwXA+hFbsnivDMOWH7aKMW6BmSX8lrQ2UummUjBVXtYaiubvrOZZF9XoGvPFJIvS5GoSEOydHmdw/ZKG57uwkoP/lLFzM0Ct2C9FYZHkhgSIXPZdvZ8woHEvpfmUhZQQgZ70XFt9hGA+ALTWMW6BmSX+lbQvUGkvp3N31HEvi6sVC10lkjAlx1exrUlkRKJ6l5LMDLmXo7R6rZUvb7fK6PJbS/RiGcV9Y1DBugZol/ZW2LVBrLKVzd9dzLI7fvw3BqD5VsPk7cqE4nyphAaWcHUQAr71j7dtnu11el8dSul+ZaE3cIoFhXAS7nxi3QM2S/kpbG6g5lpK5u+s5FuW/3QGiwJ7RoEHihYp9RFA8DWU8UQfu3jWUZzxVjwgQQcFDr1ALhsKzG6YB2sWH2pWRRGNRsO+C+VkWywsJlQXzzUHA8lhK9yuL4J4X4VctCzlbqsJwouQsDYZh3CClm/j111/FkpiYCJOJfpUU8tFHH8Fd4ZRu9yYzPbWgpL8KmUEXzhS0LVAh+0nBWErn7q7nKHP5rA6TB1RGZpoOt/TMxpCEZGjLYc+tSICwTKbeKIKC1ch+orEo2LesrCal+5VG6IOnoKuUh+yfasBwLsyueTIMU7GUbrtFzdSpUzFt2jSR1l21alVoipQUXbfOtteNO8GixjNQs4VAWW0L1B5L6dzd8Rwz0zWYMrAyLp32R/1mery2NBEBQSh3gT2qR7Ni5tYKF9ijAF2KZ1GjTo3SsSpyzLCnjkIbZELml3VgSinHBWQYxnmihoTMO++8I1oneBosahimeAz5wNtDK+Hfv4IQW8WANz5ORFQlWyus0gJ7ZJ3ZnTHIrgJ7XoPOhIhnj4q76StuAPRcp4ZhnClq7DYs6/V6dOzYsaLzYxjGjdj6Q4gQNEEhJoyZm1QuQUMWGhmy0Ph0I0u9FqCFYRinYvd/HTW1/OyzzxwzG4ZhXELX+7Px4AvpGDw9GbUamr+YGfvRhMk9nyjAmgvvMYyzUZTaIDeOJCgweOnSpfjll1/QokUL+PvbNmybPds7G9QxjJJ+TWr2dHImFBr30Ivprp6G27DnYiaOJuWgcaUQtKoaar+lxgGNLMs7J4bxJRSJmn/+kQt9mWnVqpW4PXjwoGNmxTBuhJJ+TWr2dHIWZ4/648dVYXh6fCoCg9Wpq0Lp21gFj4WCg8f/chqpubL7LQlRQVq807MO4gvq2JSGlO2H/NNhMF4NUndOv55Gao7VnIK1eOd2ZXNiGF+iXCndngoHCjPl4a3BA0R/JpNpfkGN2y3QaoehWftWGL9opeJ93InUa1pMfqoyrl32Q+9+GRgwxv6u247MeHIVA9YdRb7OhOHDC+vUzJsH+Bu1WPlAQ9fM6ZujyNcWMyeTFiv7uGZODOM1gcLPPPMMMjIyrtuelZUlHmMYb0Lu12QWK9R9qGZBv6Z5YjulZivZx53Q5wKzR8UKQVO1dj76DqqY24kynu4+fUEIGsp48lRBQ+4dstCQeKCeT5Urm29pnbbvvZTlmjnllDCnHNfMiWHcGbtFzcqVK5GTk3Pddtr28ccfqzUvhnELlPRrUrOnk6Mhu+yS12Nw/GAgQiNMGD0vCWERUrmtM53v6WApsOfpKdwUr1Ja76fD17LLHsSf+mNJ7jUnhvEhtPa4bsjsQ94qstTQurykpKTgxx9/RGX6GcEwXoSSfk1q9nRyNF+9H4EdG0Kg85MwclYSqtYyZ+uUt2JwtVV9y10x2N1oGBtcau8nCtAtHQnhTx5H+NNHoQnTu8mcGMa3UNzYJSoqSlQPpqVhw+v9uLSdqg0zjDehtF+TWj2dHMmfPwfj66VmX/QzE1NwY7s8V0/JrWhTLUwEBc+bZw7ItY5foe1lZhwFmqDxN1tppBw/9eYUXMKcghXMiWF8DMX/eZs2bRJWmu7du+Orr75CTEyM5bGAgADUrl0b1apVc9Q8GcZlUAaTuV9TYRVtEiu03Z59XE1sFSMioo3ocm82buvDbovioCynsRtPY8aMwuKDcvZTWWhDzencphwdYFSv8B5lOY39pcicCrKfGIYpp6jp2tUcH3Dq1CnUqlXrup5PDOOtUEo2ZTCV1q9JyT6uplErPRJWX0FkrP3Vgn0FSpGmLCcKwKV4FXtqwhTWqPFTf059yjcnhvE1FP337d+/32b9wIEDJe5LBfkYxhshkVKWUFGyjzPJztQg6bIONRuYY2ei49QRNHIHburx1LuGZwcIFweJBnuFQ2E1YX+3mRPD+BqKRA0V2yPLDLmfyrLQGI0U/c8wjKsxGoAFE2JweE8ghr+TjFadcis8plyP5lFLgLD3CZryIltqTFmOETUMw5SNIscvuZxOnjwpbimepm7dunjvvfdEpWFa6H79+vXFY4xvQrVa9m7bpEpNFqVj7f9zM75eOhf7dxTNOnL/c3QGn86JxL5twSKNm2JpKgrVo5EL7KmV8XQhXY/dFzNxMUOdbCFn15BZfeCqpVaMNtRQrPvJ2eeo5HhK56Tm3D35tWa8zFJDQcAyDz/8MObPn4+77rrLxuVUs2ZNvPbaa+jTp49jZsq4JWq2B1A6FsWsvPbUg8hMu2rZFhYZhzc+XYf46oXvVbXwxBYIG78IxfrPzVWzX34jBfVuLH+TSjl9W7iblgwqKLBXsfll5Bkx98+L+PtyYfG4dlVCMbJTNYQF6OCJrRSW3hCNKP8wGJOCLOc4Z8dF7L5YeI5tq4ViVAfHnKOS4ymdk5qvj7OvA+Pb2B2iT/E0ZKkpCm07dOiQWvNiPAT6sqf2AMCn1E1I3NI6ZQI5aiyzoNHb7Efrrz35gIpnZv+83IUDOwKxcqZZbD06JA0397i+WGZ5oRYIakBfmCeuZFldUYj1Odsuwt0hQUOtFCZOBNasgbil9RfeSUHOhhowXjDHvdAX+dH0LJv9aH32dseco5LjKZ2Tmq+Ps68D49vYHabfpEkTJCQkYNmyZSKVm9Dr9WIbPcb4DnJ7APNHHrUHQEF7AEmkNpObRmnQrNKxyOVkttDY7keFzzLT+gtXVItburjlOTqDCyf9MG9sLExGDTrfk4X7nr6+pYmrITcEWQCuewUloP/lLOGeqBYe4NatFOiLmdoVEPItpVyTK4qCec2ulqwS9lP/HJUcj66vkjmp+fo4+zowjN2WmiVLlmD9+vWoUaMGbr/9drHQfdpGjzG+g5rtAZSOdfzgP6Xud3z/HsXHVHNe7gJ13c7O1KJRqzw892oK3LHywuVMfSlX1OzecVdKa1sQHk5tC7JszrGk9gZqn6OS4ymdk5qvj7OvA8PYLWpuvvlmETT85ptvilgaWqZPny620WOM76BmewClYzVo1rrU/Rq0aKP4mGrOy114ZkIqHnwhHSPfTYK/ij+Av5mYqtpYVcICSrmiQFU3/uVeUtuC06eB774DnpycbnOOJbU3UPsclRxP6ZzUfH2cfR0YplxVokJDQ/HCCy+oPxvGK1sIqDlWi45dRVBwZtrggsaB5v2AIWK7mq4ntc/RUZBbgCCrjM4PeOjFinXdLprxRIKGMp4SJm9VJeOpekSACDodeiVLzF1+BYdpgHbxoW7tjiiplcLff9MPPiDA6A9DwTlSMOyC+Vk2+y2Ybw6SVfsclR5PyT5qvj7Ovg4Mo5Go+EwZfPfdd7jzzjvh7+8v7pfGfffdB3eFmm9GRkZi2ZZDCAkzZ4YwFSMzPbWgPYAK2U8Kx7py4YwICnZa9pOK5+gIvv0oHNcu6zBgTCr8VCqRYpPxlGHOeFKTTL1RBJ16YvbTlUy9aKVQmP0EPNVPg6dfkJB/PBw5v1W3nCMFwzor60fJ8ZTOSc3Xx9nXgfFOsvON6Lf2mGisHRFh7mFXblGj1Wpx+fJl0YWb7pc4mEbj1sX3WNQ4DjXbAygdi4KCKYaGXE5qW2gqMi9n8tevwZg7JlbcH/nuNdx0m3oF9pzRgZsCRSmugtwQnvar3bptQfsHMxHYPAV5e2OQ91dll56jkuMpnZOac/fk15rxMlHjLbCoYbyJk4f8Me25OOhztejdLwMDxqSpaqFxtKDxJoJ7nod/3UzkbItH/r/Rrp4Ow/isqLE7UDg3t+K/BBmGqRhJV3SYNaKSEDQtO+XgyZEVFzRM+SmpmjDDMM7F7v/AqKgokeVEXbu7deuGjh07IjjYnBHAMGrVhqFUalVcWQrGoto3lCruLDdWRcnN1mDWiFikXtOhRv18DE1IFgHCjHOhGiyUskzulIZy3ycHNbMkNh5PwYHEbLSMD0WP+lFuMxbDuBN2fxT+8ssv2LJlC37//XfMmTMHBoMB7dq1s4icnnJlJYZx85YLzm63oBZLpkTjzJEA0c9pzLxrCAnzGQ+yW1Bc2f9pzfzQ7oZQSA4QNceTcjDh1zPQF4Qrbj6TgSW7L+Pt22ujXkyw/WP9dgZ6g9VYey7j7R72j8Uw7kiFYmpI0OzatQvvv/8+Vq1aBZPJxIHCTLl5a/AA0X7AZJpfUPprC7TaYWjWvhXGL1qp+lgv3NamoN3CIss+wGCERQZg6SZ1i/ipyT9/BOH916MxanYSGrZUr3iZozOevIVpm8+JMv9Dh9mmKDeMCMXkrjVVP97DXxyGfxAwfHjh8ebNA/JzgS8faWzfWF8ehn9gMWPlAV8+bN9YDOOOMTXlMlofPXpUWGrkJS8vD/fcc4+w1DCMJ7RcuHr+rFPbLahJ6865mPv9ZQQFq2ehkTOe/rJkPLGgcYey/+QmIgvN6OHFHQ/49USqYveRGMsAjB5b8bEYxl2xO1C4evXquOWWW/Dzzz+L259++gnXrl3DunXrMJzkP8N4QMsFZ7dbqChH9wXgyvnCmh5qChoqsEeChgrsfb90PWc8laPsf6tWjin7T3EvxR1PXt93JcslYzGM14iauLg4ZGdni7o1tFy5cgU5Oep1AS6LRYsWoU6dOggKCkL79u3x119/Oe3YjPe0XHB2u4WKcOmsnwgMnvxUZZw77qeqdYYEDbmbSNCwu6n8Zf/j4oBXXlG/7H/zyiHFHk9ep0BfV4zFMF4javbu3SvEzPjx44XbaeLEiahUqZLIgpo0aRIcyZo1azBq1ChMmTIFe/bsQcuWLdG7d28kJiY69LiM45HbEVDci9kldE7cUjsC2l6elguljSW3W6AYGut9HNVuobxkpmswa3gsMtN0qFzDiPgaBRGearmbCurRsKCxt+w/sHEjQB8958+bHwvQ+6teVK5ng2hQ0V2Ke5GPR7e0TtvtcReJsfxKGMvPvrEYxisDhZOSkkRMzbfffovPP//c4YHCZJm56aabsHDhQrFOx6tZsyaGDh0qRFZZcKCwe+PslgvObrdgL4Z84O0hlfDvriDEVjHgjY8TEVWpsDS/Gv2cWMzYT9Gy/wkJwC23AKm/Vob2RIzqxzuZnINxvxRmPxEkaMqT/STGokwqK21MgoaznxifrSj89ddfWwKEDx06hJiYGNx6660iSJjSusl64gj0ej1CQkKwdu1a9OnTx7J9wIABSE1NFcKqKGRJosVa1JAIYlHj3ji75YKz2y0ogf4rP5wehd++DkNQiAmvf3QVtRqaa6GUF64WrC5y2f+OQ64gsFI+sn6oCeMFx7lwKJCX4l7UqC2j5lgM49HZTy+++CK6dOkiunSTiGnevDmcAQUjkxUoPj7eZjutHz58uNjnJCQkYOrUqU6ZH6MeJD7U6q2kZCwSMu4iZmR+WhUmBI1GI2HIjGQWNG4IuZpoCYi4KNalLMdWQCTxoZYAUXMshnEn7P4v9KT4lQkTJogYnKKWGoZxZ4wGYMdGsyvgiZFpaNOFW5O4Lf5GaAJMDq8mzDCMMjymuDoFI+t0OpFtZQ2tV6lSpdjnBAYGioVhPAlqefDq+1ex7ecQdLvfnIbLuCfasIKeT7lawGB33gXDMCrjMf+FAQEBaNu2LX799VfLNgoUpvUOHTq4dG7eDhWz27ttk4hPqcg+roD6On29dK6Im6no3B19jvlWJU4CgoDb+mRDoyn7eWrOa8/FTKw+cBV7L2WVUYAuU8SUuNPxlIylNpJBA/2/UdAfiyy16N3sPy+IOJaKnKPS81Pz9VGC0uM5e17ePnfGwy01BLmSKDCYek1RU825c+ciKysLTz/9tKun5pUo6Z+kZr8mNVHS10np3J1xjvpc4M0X4tC8Qx76DkqHVsHPDZrX4olD8c92c20donWHrngpYWGx86KMJ6wqfiwKeJ2w4TRS9HJ2VRKiA7R4u3cdxBfUZqGeR3P/vIi/Lxd+qbarEoqRnaohjNJx7ICON/6X00jNLTxeVJAW7/S0PV7RHkuUTj2qg+3xlIzlKKSMAORuq1LuPktKrqnS81N6vdRC6fGcPS9vnzvjJZYa4tFHH8WsWbMwefJktGrVStTMocrGRYOHGXWgL3Lqn2Su33JW3NI6pUrbs48rMAsavc28aJ1SuO2du6PP0WSiJpUxOH4wEBu/CEXqNWX/liRoTu/cajUriPXFE4YUGyRcbVVf0dOpOEjQmPQmm7Fofdz6wkrO9OV74kqWzT60PmebOVDWHuhLOl9nEu0G1qyBuKX1sRsLj0dfJtRjyXofWqd0anvHcgUkaKjPkvW8aJ1Squ25pkrPT+n1Ugulx3P2vLx97owXWWqIIUOGiIVxLEr6J1HesVr9mtR2OZXV1ymuag1Fc1ezJ1VJfP1+BHZsDIHOT8LIWUmIqVx2LRqaF1lorjtDkxH9t2+2zEsWNKVZacitQRaa668W0F9vEu6OuFB/YU24bh8J6H/Zvp5HdDyyOhTfP6nweEp6LCkZq1VVx6VZa0LzIVEsTR4JUY1dfZYax4WUeU0vZ+gVnZ+ze1IpPZ6z5+Xtc2dUEjUPPvigXXVsGF/rxVTyPq4QNUr6OpksRSJLn7uS61CRc9z6YzC+/sBcc+HZSSlo0laZz16eV/GzAtJTVuCGlo0tgka20qyYufW6dO6jSTmljnX4WjaMUnCp+1yy4wNePl5JPYisj1fSPvLxlIzlSFET3O0S/KpnI/vXqjCciLSrz1JUsF+Z1/SYwvMrqSdV0eulFkqP5+x5efvcmbJRZOemKrxKF8Y7UNI/Sc1+TWqipK+T0rk78hypSeXSqeYKtPcOyLAr00meV/GzolYRsYUWmjJoGBtc6liNK4VYeh6VeDw7Ptzl45XUg8j6eCXtIx9PyViORBtmrh8kZfnb3WdJyTVVen5Kr5daKD2es+fl7XNnVLLULF++XMlujBch9086uHOYcLWYfz9uFv2TmrUv7MWkZB9nI/d1ykyjvk6F8yra10nJ3JVeB3vJztBg9iuxMORr0O62HDw6NM2u59O8KCh46M6twuUkn+EwrQ6t29+KanXjzEEaBVx84iu0JffTGLLWmLfJFps21cJEUPBgvanI1YLYLls6KIB16JUs4R6xHE8DtIsPtesXKx2PAl3nzTNZfvnSFwX1IKLt8vHMPZaybPahnku0XT6e0rEcgwRNqDkK2JTpd12fpaX/XBHzKDov6z5LZV1TWpScX2FPqtKvl1ooPZ6z5+Xtc2cc3PvJ0+DeT1C9f5Ka/ZrURElfJ6Vzd9Q5/vF9CH5ZG4oJi68hKNj+f0OaFwUFF5f9VKX2T8VaaqyDha1dUVcy9SIouDD7CddlP1HPIwpgVSP7iY5Hga6FGT24LqOnaI+lkjJPlIzlCDTBBoT3Py4EScayRoCksbvPkpJrqvT8lF4vtVB6PGfPy9vn7qtkO6r3E0H9l7744gucPXtW9GSyhrpnuyssahzXP0nNfk1qoqSvk9K5O+IcKfNJSfp2aRQ3L5sA4WIoKcaGAk8pToPcGiVZOeSeR2R+r+ivVTWPp2QsNdFWykHYg2dgyvJD5qoGFeqzpOQclZ6fmq+PEpQez9nz8va5+xrZjhI18+fPx6RJkzBw4EAsXbpU1Ig5ceIEdu3ahcGDB2P69OlwV1jUMK7mp8/C0KFXtirdtkuiLEEjU1rwMFM2fnUyENLrAgxXgpD9rWvixxjGV8hWKGrs/o343nvvCTGzYMECUeV37Nix2LhxI4YNGyYOxjBM8WxYE4pPZkVhysDKyM1WUCq4HGLm3f13Kg4QZhwTJMwwjAfVqSGXU8eOHcX94OBgZGRkiPv9+/fHLbfcgoULF6o/S8btoboplGbsbu4nd2Hfn4H4eJbZ9dCjbxaCQiSXWGfKgmpzUCpraWZ2JfuoidLjqTl3JftdOaeDbmcodKmBqKizy9nX1F3h68A4XdRQ88jk5GTUrl0btWrVwo4dO9CyZUucOnUKPhRzzBTgrm0S3InzJ/0wf3wsTEYNutybhXsHmn8IuJOgUVKuX802CWrNSe25KymLf/1YWWhXJadc18HZ19Rd4XYEjFrY7X7q3r07vvvuO3Gf4mlGjhyJnj17ihYGDzxQWIKe8Q3ctU2Cu5CeosXM4ZWQk6lF49Z5osCekiaVznQ3DRxzK179zb/Mcv1qtklQgtLjKdlP6VhKyuKreR2cfU3dFW5HwLjMUkPxNNQdm6DA4NjYWPz555+47777MGhQ8X1lGO/EGS0EPBnquk21aK5e8EPlGgaMmJUE/wD3sc5Q7Zqq+AqZb3XH6dRDpZbrp/tqtUlQ6oZQcjwl+ymdu5Ky+NZj9akDBCQBj2eU7zooPUdvh9sRMC4VNVqtViwyjz32mFgY38PRLQQ8nYxULTJTtQgJM2HM3GuIiDa5VfyMzK4RP5RZrl9GjTYJSpBL1Jd1PCX7KZ27krL4Mp11wK5lAHRAh75A12TbsdQ8R2+H2xEwLm9omZKSgg8//BD//fefWL/xxhuFKyomxlzynfENbFsIyL81Xd8mwV2gxpRTVybi0ml/VK9nVYGtgmKGUDO7qbafXymvorkcvBwuV9o+amLdQqC04ynZT+ncrcviy5YCeb3oWH9WAqroAE0+EJBSvuug9By9HSXXnWEcJmq2bNkiXE2UJ96uXTtL7Zpp06bhf//7H7p0Kb7AGeN9OKqFgDfE0chWmdBwCQ2aK2tS6UzrjDX1/PzRPTAQw/LybNokDIUW7aoEW34lq9UmQQlUol7J8ZTup3QsJWXxaaz5cVmYQYaaq8AqqXzXQencvR1uR8Coid3F95o3b44OHTpg8eLF0OnMUelGoxEvv/yyiK05cOAA3BUuvqc+7tomwVWcPOSPN1+IwyOD09H7scwKBwWTmCEcXXsm1WTC8JQk/JaXZ9nWqXUnNAp8CEH+oaI4n5ptEpSg9HhK9rNnrLLK4tM+m6SzeGJoHvbuBUaOLP91cPY1dVe4HQHjsorCVJtm7969aNSokc32I0eOoFWrVsjJyYG7wqLGcbhrmwRnknRZh9eeqozUazq07JSDMfOSKtQCwVHuptI4ZcjHaYMBdfz80GmNOVauaNVhTy7Dr9ZYAS2TENT+Ki7vC0HqL1UqfB24DL8Zvg5MRUWN3e6nNm3aiFiaoqKGtlG9GsY3ISHjq2KGoArBs0bGCkFTs0E+hiYkqyJonF0ZuK6fv1jk7CjCurs3CRu5e7SzUHo8JfupNZZcTTjaFIwQFa6Fs6+pu8LXgakodosaaocwfPhwHD9+XFQQJqgA36JFi/DWW29h//79ln1btGhR4QkyjLtjMgKLJsXgzJEARMQYMXruNYSESW7tbmIqhibMHPhtyixXrgXDMA7C7v/Ifv36iVvq+VTcYxqNRlQWpluKtWEYb2f1gkjs3hwM/wAJo95NQlw1o8e4m4py0pCPMwXuJ9liU5Q9FzNxNCmn1G7Rrih3r+a8ytrPcDIcVy5L+Gt7LirlZJXaNdtd20p4MnxNGdVEDbVDYBjGzOF/AvD9x+b4rEFTk9Gwpd6j3E0yKRQonJaETTmFgcK3BQdifmQsqhWsp+UkYuBXR5Gil+vtJCE6QIu3e9dBfEFarivK3VMMxoQNp1WZl5L96HjjZ15Bao58vDREBWvxzu2Fx3PnthKejLPfX9y+wfOw2+tPPZ+ULgzj7TRqpUf/V1Lx0Etp6NjbfYPky4IEzW5tnk2ZeloflpZk2eebvQkw6U02Jf1pfdz60y4td0+CRq15Kdlv/K+nka812exD62N/KTyeO7eV8GSc/f7i9g2eR7kcwp988gmWLFkirDbbt28XAmbu3LmoW7cu7r//fvVnyTBuCqVs3/lEJjwZcjmRhab4MvV52PboamS+cgZZhmJK+gPorzdh76UsxIX6O73cPbmcyEKjxryUlOu/nKFHHkx4bRLQoUPRfczHI1eUs1sg+ELLBWe3U+D2DT5iqaH6NKNGjcJdd92F1NRUS9xMVFSUEDYM4+1kpmvw4YwoZGeo1JnSxVAMTWll6inF+9c3fi21pP/ha9l2tRlQC4qhUWteSvaj4zVpUihoiu5Dx7MeS0n7BjVw9vFcgbPfX654PzMuEDULFizABx98gEmTJlmK7xFUXdidC+8xjBoY8oF5Y2Lx69owLJioTlsQOePJVchtEuSy9DLyOgUNtw4oLOmPYkr6U3Cudbn74sZxRLn7hrHBqs1LyX50vPj46+ch70PHsx6rpHk5sq2EM47nCpz9/nLF+5lxUaBw69atr9seGBiIrKzCYCqG8TbIlL/i7Sj8uysIQSEm9BuaXuEx3SHjidokUFDwwnl5NmXqF84zBwvLtWviNBoMliSbVgpDqBWEX6gl+8fZ5e7bVAsTQcGD9abr5kXb7ZmXknL9tFwKIQudhPPnAdJ6tM+8eRDBwvLxnN0CwRdaLji7nQK3b/BM7K4oTM0rExISROxMeHg49u3bh3r16gkLzvLly7Fnzx64K1xRmKkIP34ahk9nR0GjlTB6ThJad86t0Hiuzngq2iZhWAnZT1EFVQTJTfXA1Su4avWRER8RhCWzvsXGT85YWik4u9z9lUy9CAouzH7CddlPSuelZD+pwwVENs/ARx9RfKF5n+Kyn9y1rYQn4+z3F7dv8IE2CcuWLcPrr7+Od999F88++6xYP3HihBA6dP+xx8yl1d0RFjVMedm9OQizR8VCkjQi26kiwcHuXGDPuk1CSXVqtuTmYI9ejzYBAegSFIxqq/qKVgqE3E7BFeXuKUiXYlpKq1OjRpuEkLvOwq9GNo58E4PNv2pKPZ49x1QLX2g1wNfU98h2lKghVq1aJYQNiRmiWrVqmDp1qhA57gyLGqY8nDnqj9efjkNejhY9+mbimYmp5W5U6Q7uJkdAwqa4PlHeSOjDJ6GL1iPr+5owXixZzDAM4wG9n4gnnnhCLNnZ2cjMzETlypUrMleGcXtCI0xo0FyPAWMrLmi8SczIlNQnyvuQLH2fpMziLVkMw7gOu0UNdeEm405ISIhYrl69KlK5KdamV69ejpklw7iQ2g3z8cbHifAPlFCCR8Zud5OSdgSuoKLzInFDwuajWsC0/RHQpVfHpI7/eVyJ+hKPqQHy9lSCNjQf5y+bcCk1k10SDOPJooYChB988EG8+OKLok7NzTffjICAAFy7dg2zZ8/GSy+95JiZMowTMZmAS6f9UL2euYZLdFxhEGpF3E3UjmBEShJ+yysMyO0eGIh50YUBua5ArXlty83Fkz3mwWB1uR48B0zrVgPN4sPcvux/mWXxJQ2S/ooq2KewgjAHjzKMe2D3pyhlN3Xu3FncX7t2LapUqYIzZ87g448/xvz58x0xR4ZxOl+9H4GJj8fjz5/NdVAq6m6SLTQkHPbm5dmUsqf14SmF7QhcgVrzejL1GgKDYVNWPigYmPz7eY8o+6+kLD6XzmcYL7LUUBwNpXITGzZsEFYbrVaLW265RYgbhvF0tv4YjHUfmAPR9Hka1bKbyLVDlpBiS/rn5YnMI1e4otSa14L0NGGhGTu8uLLywFf/XkPfppXctuy/krL4fuH5SA3OwsjngNtuK34fdkUxjAdZaho0aIBvvvkG586dw/r16y1xNImJiaVGJDOMJ3BkbwCWTjVXCr53YDq63W8ue2+vdaZ1p/nXBQTL7QhKKmVPqdSuQK15bS1wXZVUVn6PlRvJHcv+KymL739jKpYtA9q3L3kfhmE8SNRMnjwZo0ePRp06ddC+fXt0KGiCQlab4ioNM4ynkHhBhzmvxMKQr0G723Lw6JB0VbOb5HYEJZWyp+BcV6DWvG4NDCy1rHybKqFuXfZfSVn8qBhzBQyqJlzSPgzDuA67P0Ufeugh3Hrrrbh06RJaWv2k6dGjBx544AG158cwToGaU84cXgnpKTrUaazHy28mQ2l8rNJietSOgIJvh+Xl2ZT0H14QlOuqLCi15jU0IhKzszNEywDrsvK07qcDwiJnAJjttmX/lZTFD4k2N/D9Zh3Qug2XzmcYd6Ncxfc8FS6+x5TEdyvCsXp+JKLjjHjjkyuIqWxySO0Zakcw3A2zn9Sa15+5uXgi9ZpN9hMJmo2z+iK0/ht2FeZzRdn/ssrih/U7Dm24AQumBOHrLYVtMjj7iWE8uKKwp8Kihikthfvr9yPQpmsO6t1oLq7myGJ6StoRuAK15rUoPQ1b8vLQJTAQgyMixTa5nQIJG0KpuHFFifpij6mREP7sEWi0QMaq+rhwWeLS+QzjJLxO1EyfPh0//PAD9u7dK+riUI0ce2FRw6iBO/ducneK6xPlKWhC8hH+5AlIJiDjw0aiZg3DMO4lalxn77YTvV6Phx9+mIv7Maqx789AvPdqNPSFHpcKZTcxCisOh78vloFjbsX686PgKWjDzFlgUpYfCxqGcVNck25RDqhhJrFixQpXT4XxAs6f9MP88bHIydSiWl0D+jyb4dO9m1zdJ4pwd6uNKcsPuTvjABMLGoZxVzxG1JSHvLw8sVi7nxgmPUUrMp1I0DRuk4d7nsrweneTmr2mlIylZB+5TxQJG6KiTTAd3SNKyvKHfl+s6uMy7o8r+o8x5cOrRU1CQoLFwsMwRL4emP1KLK5e8EN8DQNGzExS1KTSUwWNmr2mlIxl7/FI2FTFVyLWprzdvcvs18Qw5YTfW56HS2Nqxo8fD41GU+py+PDhco8/YcIEEVQkL1QFmfFdKCT+g2nROLo3ECFhJoyedw0R0eVrVOkpqNlrSslYahzP3jgbZ/Vi0lbKgTY2F/Dz7vcMUwj3+fI8XGqpeeWVVzBw4MBS96lXr165xw8MDBQLwxDffhSOrT+GQquTMHxmEqrXdU1bAmehZq8pJWPRfWf3tlLSr0ktd0HwrVegq5yL7PXVYTjD2ZPejjPfW4yXiJq4uDixMIwzaNQqD2GRRjwyOB3N29uR8uShKOnppFRk2NMfSo3jqdmvSa0vHk2YuX6RKdN96goxjsOZ7y1GPTwmpfvs2bOiRg3dGo1GcZ+WzMxMV0+N8RCatNVj1tdXcPtDyhsrenLGk5q9ppSM5YreVkr6NamC1gRtiLlFgpTp1aGIjLPfW4yqeMx/JzXSXLlypWVdbp65adMmdOvWzYUzY9yZpMs65GRpUKO+2ZKgNIZGFjOEJwoatXtNKR2rvMcrbyaUkn5NaqAJLahRY9BAyuMAUV/AWe8tRl08pqKwGnBFYd8iN1uDqc/EIfGCH16Zk4Qb2+V5vXXGkb2mlIylxvGs2ykoETZl9WtSA13VLITeew7G1ABkfVH+OD/Gs3DGe4tRt6Kwx1hqGMYeTEZg0aQYnDkagIgYI+KqGXxO0BAkJFbGxqnS00nJWGocz9pqo6QwH325TO5a06E9oizVhNn15FM4473FqAv/hzJeyeoFkdi9ORj+ARJemZ2EuGrmeAhvdjeVBgkLtYJ0lYxV0eOVxx1VzYFfOJYg4SwOEvZFHPneYtSFRQ3jdWz6JgTff2x2Lw6amowbWpizGHzFOuMtqFGYTy0M50ORa9TAlMIlIhjGnWFRw3gVh/4OxEczosX9B19IR8feOWUKmglrPsC4d5JxykktBFyB0nmp1QJBTex1R6nBnouZOJqUg8aVQtCqaihMV4OhvxoMV8Kl+hmmbFjUMF7FhjWhMBo06NA7G30HpZcqZl774SOc+eA8Uvdl4qmC7bcFB2J+pGNaCLgCpfNS2gJheFoSNuUU7lPe61VeYdN2GgqCiB0jbChuYvwvp5GaK2fJJSEqSIt3etZBfEGKr7PhUv0M44V1ahhGCUNmJOOxoWkYNCUZGk3p7qaYMadgOp5pUwJ9tzYPw9Ic00LAFSidl5L9SNDQ9VHjepVX2AhxE/4+Bo651e52CkogQZOvM9mcI62vST1lbpGgcX6yKJfqZxjlsKWG8XhMJkA2FJA35L6nM8rstj177FV0zckroQR6Hk6Fq9tCwBWuKKXzUtoCYZNK18td3VHkciILTdFz1OmA7t3pCpxG+kcNAUMJatkBcKl+hoFo3KczKKsxxqKG8WioytKH06MQECThyZFp0Pkpy27aVFDWv6QS6Gq3EHCFqFE6L3taIKhxvdzVHUUxNKWdY26WBjA417jNpfoZr0aS4GcwISjHgKBsg/k2xwB9oA5n60dZ9rnzq2M4F6UsSJ9FDePR/LgqDJvWhUGjldChVw4attQrym6SS/pThVD5l6+8XpEWArKlw9HtAdScl5L9ZKeLGtdLTWFDWFttKiJsGsYGF3uOZ84AsbGAIcPPpaX6i7vuXKqfcVe0JFZyC8WKUafBlRqFRW+7/nQKoZl6+Bmud+kmxwYVihqNBvoAHQJyldUaY1HDeCy7NwfhszmR4v6To9KuEzTW7qai6dpU9p+CXBfOy7Mpgb5wnjn41REtBJyN0nkp3U+t6+VIKM6mvMKmTbUwERQ8b57J5hwPHwbatAEC8wJQei6d+nCpfsbd0JgkBBZYVMgRm1KpMCvw5i3nEZKpF48F6E3XCRVrUeOXb7IImnx/LXKD/SxLRqStVWZrz9pIo3i2dSfKnB+LGsYjOX3EHwsnxkCSNOjxUCbu6JdpdzE9ytqhIFeKCSmazWMvlCVEQbUUg1I0e8iVKJ2Xkv3UvF7uCmU5jd14GjNmFH4gv/S8+VZyUXduynKioGCKoSma/cQwqmGSEJhnFLEr2VZiudnflxGalW+xuATkGYWYIVJig/BHrzqWfcNT88S+MmSdkYVKehH30a5bq8NIYibIT9yWRl6wH0z5pRdQlWFRw3gcKVe1mDUiFnk5WjRrn4sBY1JtMp2UFtOjNOSPo+NEkKszWgi4AqXzUtoCQa3r5a5Q2vbKBxpi76UsHL6WLerU3NIujTrHuayaMJfqZyoeZCvZCIf6/yUhJDPfEsMillyDSO5LiQnCH70LhUr8pSyEZhYKFcKkhVmsBNlKiAPt4mHSFgoZAx2zhDTU9JggOAIWNYzHZTrNeSUWyVf8UK1OPoa/kyQynjyphYArUDovZ7RA8ASo4B4thDbUnLJucnHfJy7Vz5RGtTPpCMmyFSqym4isJNZCpc7x1OuECiFpRJdrm21Hm8YKsWPtHqJA3uLESmK1MLgaFjWMR0Gp2/c+nYFPZuowZt41hIb7TJN5xkWF+fT7Y2A4FwpTomsrCjO+R8zVbARnkTDJtxEptJC4+PP22pZ9G++/irBihApB+1tzpn6UcDNZCxUhVoL8IGltxcq5egUBux4CixrG47jptly06nQZ/ir9aHXX1gZqsjk3B//o9WgTEIAuQd735ax2JpQ1hnNhAC0MowJBWfkIFsIk//pU5gAddt9a3bJvq52XEJZRvFDRGW1/0F2pHoaUPCNyQvyFQMmzFixF3ETHb/SeOLiisKhhPALKdKp1Q76l23ZJgkbOePLk1gZqQrEvD169gqtWJuU4jQbr4uItqdzehG1371uxYuZWlzXBZHwsI4jSl4tx/VCw7MF2VSz7dtx0DmEZxTfZJQFiTUpscIEwMQuVoos1/7aJd9DZeRbe96nGeB1H9gZg3thYhISb8MbHiRZhUxQlGU8ltQboUlCnhdKaKQuIgma9ARI0ekmyOcfBkoQHrl7BnqqFvwi9Utig4lYbTbABuso5MKUHcIduX0SSRLaPTUBtjkGUPjjWrJJlt24/nUJ4eglCJUhnI2qyQ/2FCLKxpBQsOSG2X8n/cIab3bCoYdyaK+d1mD0qFoZ8DRq11CO2SumCRomYcefWBmq7nMhCU+w5ShK25OZ4pStKTXeULj4HIb0uwJgYhKxvCgMtGS+oZJtfUMm2YNGapMKCbwA6/HoWsdeyoTUVL1SsRQ25d0Iz9MgLKhAoIdZixfZzZMdtNR17bj4OixrGbcnO0GDWiErISNWhTmM9Xnoz2dLjSUmBPU9sbaAmFENT2jnu0eu9VtSo5Y7ShprjGUwuqlHD2I/OIlbyRdyJdUZO6+0XEXMtR7iG/IrEpORZl+YvQBY09Ji1WMkhoUIu3YIMoF1dqsOg0wJFgmwZ58OihnFLjAZg/vhYXDjpj+g4I0bPvYagYKlC7iZPaG2gJq0DAko9Rwoa9gUq4o7ShBncIp2bAbRGs1jx05tsapw02ZuIqKRci8XF36rxIYmR9Q/eYFmnx61TmfUBWpt4FWuhsu/mKqLmCllfJF3pYsXgr1P5bJnywv+pjFvyybtR2L89CIFBJiFoYiqbKuRu8pTWBmrSNShYBAVTDI31OQ4pCBb2diuNGu4obZj5C1ByUeE9X4DiSyhuhWJMZOoeTUZkcp5N0G2g3lisUIlKzkVcYrbNmAY/KgDnb45RsRIq/7WIwxFJsmQEmfxKTgiwrqrLeA4sahi3IzdHg6P7zB8oL7+ZjLpN8ivsbvKk1gZqQllOFBRMMTRFs598EXvdUZpQttSUGysxQVQ9m46ItDybWitCsOQaRTE3a6FS5Xwm4q7YChWCMomoSi0JIbmeyolGMThbL9Im4NZYguUk1apPEeOd8H8q43aQm2nyh1dxYHsgbuqeq4q7yZNaG6gJudkoy4mCgvd4cZ0aR7mjZEsNx9QUT/TVbJGeXDQ7SDQ7lIANfRrYVLEtTqgQ/vlGaIySxc1zrm4krsaHFmYIFcSyUOPDopVsE6tzDSGmEBY1jNuQm61BUIhkETbFCRo1xIwvlv0nIePrYqYsdxRhI240EjQhBpc2s3QF1GU5mArEFRSFs7as6IwmbLmjrmXfRgevofLl4oWKHAdjogBaymSsFoassIBCi0pIyWX3z9eNdPBZMt4KixrGLUhP0WLygMrocm8WHnguw/L5pqa7iWHsQgPk/FZNWGukHM8OBNUabNOXrRe/fCP+6lqYZtxi1xVUvpylSKhQcTi6UNenMZsXCrSVOdk4xsFnyTAsahg3QJ8HUYsm8bwf/vhfKO54LBMh4ZKNdeYbV0+S8T1MGhhORsDtK9kWiBMqvS/fD9Absf+mwoJvN/9xQbFQyQr3R0a2lUWlyCJZWVSOtPCOIpWM98CihnEql86cxJXzZxBfsw6q1qorYgmXvRGNo/sCRcXg0fOuXSdo3AVP7hHlyXN3BRfS9bicqUdVV3XGJrFiqWRr7hFEGULHmhYWfGu77QKqn80ocYiDrStbsntIjBh0mmKr2BYtDnfAqvotw3gaLGoYp5CZloKFE0dg//bfLNtadOiOuk0+w9YfQ6HVSRjxThIatflEPOZOgsaTe0R58tydycAx5kyojDwj5v55EX9fzsINNwBxcUDo6WAMbFQDYQE6dcru642iiaGwquQaca5eYfxIs91XUPVcBoJyzYG2RTnZKAbGAqFiKLg1aVFsfyDrcNp9N1XB3vZVrguyZRhvg0UN4xRI0BzcuRew6kJ0YMev2L/dnFr89PhU3HLHCrd0N3lyjyhPnrsrMqEeeWou0tKyxPVqcheQ3gdY+2kO5vxwEa+VVt6+SNn9a1VCLQ81+DcJVS5mCotLYI4ROpOtWrlYK9wiVHQGk3AjiSE15vL7skghKwu5m2QOt4jDf63iRGfnssRKWcXjGMZbYFHDOMXlZLbQ2HYhkqRHxb0u953H/c+5n7vJ03tEefLcXZEJFf5WCk6nHrJcrwMFmq9nIvDz5SxczNBbXFG1TqSi0uUs24Bbq7L7PzzUEEZKPwYQmqkXpfmtsZTdD/YTQkYWNSeaxOB0gyhzRhBVsi2l7L51sTqGYczwfwXjcCiGBsV2IUoH8Du63v8jpm7e6XaCxtN7RHny3F1RHO7Sy+uQAKBPwXpeZfPtrYnAcQDTU3MtoiY6KQc1ioln0ftT2X0/UXdFFjVn6keJ/kMW11ApZfczI7gTOMNUBBY1jMOJr1G74F7RLkQ/ks0A1euOAc7CLfHkHlGePHe1CNDnwT+fGhuS28cosnzk+zqTCReqVINUEFvU2GDAbVbPzS0QNYGJAOUN1bGKqblYMxwZEYE2MSxkOZEtLkWr2LqfXGcY78T7P9UYl1O1dj0RFHxgxwhIUnMA1IZgE7Ta4WjWvjuq1Y1zW1HjyT2iPHnupaEzUNPCfLNAIXFCIkUWKkYjLleuYhEqIdnZCMsuOZWZnmMo2DcgOBhf5+txwGhEjyDAUBC/2y4RyItsgnnxhe6lq1XDcLWqo8+UYRh74fQHxikMSZiH2CoLKeeJfrtSvgmatW8ltrs7lCnUKjAQ/SmWQtiWINY9oUeUp8xdYzLBLz8fgbm5CMnOQnhGOqLSUhCTnIS4q4nicZnwzAxUSk5CTFoKIjPShWgJzstFAFlkTGahI6P3D0BOUBAyQ0KRFh6B5KhoXI2phMtx8bhYpSoMusLfdTnBIWgYF48tgYF4vCCeJisLqNSwE5Yt/gjrz49y7kVhGMZu2FLDOIVdv1bDtUvNxP17ByahW59Nok6NO9aj8aYeUa6eO4kR2YJS1LKSEhVtsahEpqcptqgY/Pyg9/OHUacTi0mrtdw3anU2QiU7NFQs9l6vkw0zYUIqQpN12DK7HYDViLCjuzfDMK6BRQ3jcP7dFYiPEqLF/b6D0tB3UD1x3xMEjbf0iFJ97pJ0nduHbtPDIgqFSloqwrMySxwi3RhhESpCkGi0oqqtLE6shQotMplh4WJxJLUuBiNnAXWDBi7+9RWqrepreYwsNixsGMY9YVHDOJRLZ/wwd0wsjAYNOt6RjQdfyFC92zYDdcVKEctKdnCIRaiQWygsMxM6qdAdZE1WcKhFqJjkW43GRqjIlhX5cSIjLBwZ4e7TksAvTYfwHSGungbDMHbCooZxGJlpGswcHousdC0aNM/DC1OSER7jWdYZr0GSoLVyBeUGBlnSmUOzMhGanWUWMiaTTSVaIi8g0CJUCFnQUNisjTVFp7PpC5QZGiYWWRCVSinF4zbn5uAfvR5tAgIq3Gmc20UwjHfjEaLm9OnTeOONN/Dbb7/h8uXLqFatGp588klMmjQJAQEu6MvCKCIrQytKgVSqYsArs5MQU4UFjepIEjQFriCKNZHFQXBONkJysm0sLtaygQJlDf7mL3USMhRkaxlSFitanbnJodUTs0JCkRMUXChgShEjisRMKVAM0APJV3DNUJh1VMlPg29i4i3p6o5qF5HVxlwsL+hoAHSZnt2hm2F8CY8QNYcPH4bJZML777+PBg0a4ODBg3j++eeRlZWFWbNmuXp6TAnE1zBi2sqrMJp+wrLzc4EL7G6yN8jWWjgE5uUiKDfXNo7FWChWrIWKn8GA4Nxcm/FIGsiuH41YM5MTHIx8f/9CIUNf8iWIFRO5jXTO+ZInQZMTKGHiWKBlS2DfPmDePAl9kq/gn8rVHdou4tpjacivbkC16ZUQckhX2E5hzCDxOAcMM4x74hGi5o477hCLTL169XDkyBEsXryYRY0bknRFh9h4c1pt1bqr2DpTCv56PQL1ecUWh9NKko1QCdDrSwy8NWq1wuIiQ+4lIWCsAm5LEisGP3+xuBPkciILDQmanj3N2+TbGTMkbMnNUeyKsrddhAQJhljz+9cvyarg3hNfoSoKgoY5E4ph3BKPEDXFkZaWhpiYmFL3ycvLE4tMejqV5Wccyd+/B2H+uFg8NSYVfZ5f5pOCRhSHEzVTri8MR0tipcoWoRKUlytqrZQ4FrmVCval2Jb0sPDrUpjptqhYyQ8IEIunQjE0BFlorJHX9+j1ikWNve0iTKESpCCzJcsv+fqPSGurDQkbgsUNw7gHHilqjh8/jgULFpRppUlISMDUqVOdNi9f5/RhfyyaGANDvgbfbd+AfQ29R9CI4nAGw3VCRbasJEfHWMQHxbIoFSpUHC67IEbFJkOIUpu1Opu4FH1goFh8gdYFgoxcTrKFRl4nKGjYUe0iDJXMIkiXpoU2v3g3HLujGMY90UgShXK6hvHjx+Ptt98udZ///vsPjRs3tqxfuHABXbt2Rbdu3bBs2TK7LTU1a9bEsi2HEOLgOhe+RspVLV7rXxnJiX4Ib/on9iU/AX+N+cvBEzKCiiu5nxYRaXHLUCpzaUKFqtTmBQWJ+0E5OaLqbYnF4awCepmSaZ14QcTUDB9uHVMDBOdp7I6pGZB0VcTUzCvSLoKqKxeNqclqnYNLo5MQeMofNV+NL3NsckftziCrzVYWNgzjILLzjei39pjw0kRERLinpeaVV17BwIEDS92H4mdkLl68iNtuuw0dO3bE0qVLyxw/MDBQLIxjycvR4N2RlYSgCap6EtuTnoe/1uAWGUHWQoUEDKUYGwt+mVO9laiMtBKHoUwfWdRQlVqKWylqUZEzhCjQViY3OFgsTMWgLCcKCqYYmqLZT/ZCWU4UFEwxNEWzn4piiae5piwgmt1RDOM+uFTUxMXFiUUJZKEhQdO2bVssX74c2gqmizLqQLGpH7yZiZOHqkMXloKNxqcQ6ZfuUDeQsK5YWVQoe8dYUBqfaq5QyX0Ksi2OvMBAi6gR6coFQbZFA2rpNt8qziInJEQsjPMgtxFZZCgoeE8F69TY0y4iXxY1xcTTlAS7oxjGPfCImBoSNORuql27toijuXr1quWxKlWquHRuvs7Bv3bjz5/uhz/0+DzwOdTxO1O+gYpYVigoVk4dpporERnpZmtLMWKF3DmyqKEUaHmfopVsabHpCxQcIhZ2Bbk3JGQqWnTPnnYREVtCEHjGH/6J9n08cnYUw7gejxA1GzduFMHBtNSoUcPmMReGBPk01OqA2BD8MSaGH0Ql3TW0D/yr5LiVguJwcuArdWMOz8oojGWxSkcmrsXEIldX8EUmAf4FGSwWsVLg9ilabp8Kw12qHCi2lVn8jcUMUwwBl/zFUl7YHcUwPhoo7GwoUDgyMpIDhVUQNNM2rURkphErEjIs4iBAn4eQbHMVW0v/IKuy+0KoFPziDs7ORmxqss241mX308MjkEel/KnirdEIf0O+bSl+FiSMmyM3weQgYobxkUBhxg0hjUtLgZUj4sQR1Ni8EcFXr4glNOVf5J5KxPbUfPjBaCNUdAYjwrKzrh9SxK3ozHcK0AcEIDkq2iaF2aQpvjgcuaHynFTFlvFtJI2EtNuz4J+kQ8i+IGiMmgpZbAhrqw0LG4ZxLCxqfJCQSxcQ989fBULlskWwBF8z3/751nu40LWX2Df62H9oubD4tHuywVhXsaUMICoOd13NlWIq2VKwbrad/XsYxtEYI024NjAVMAL1B9qXNq6E9edHsbBhGAfC3ypegn96KiJPHisUKFZLyNXL2PPKFFzqdJvYt9L+v9Hx1WEljkXPkUmr1whnH2qHX3LPIDEyABlrXsRGQy8YNDp8EPcC4v0SLftSQbl0/0gHnynDOA658J5fig4aE7s4GcbTYFHjxmgMBgQlXbVYU0ISrawqV6/g0NODkXhTR7Fv1e1b0GnC4BLHCr14znI/s2YdXGnXETlx8ZYl2+p+TuXCjDJD+714MDcX30wMxKupb2ClYSCCNdn4utKDNoKGYbwBQ8z1PZ8YhvEcWNS4ApMJQSlJCC4iUmg5ed8jSGrZVuxW89cfSxUq52/rbRE12VWqIaNGLeRUshIncVWQU9l8m1bvBsvzkm9sid+WrrFryiuyBmBl9kBoYML8qGFo5v9vuU+fYdyV/EosahjGk2FRoyaShID01OuEysXOPZDa8EaxS/XNG3Dr6BdERk9xpDS60SJqSJiY/PyQU6myrVipXEXcXm3ZzvK8ay3b4fvvtjnktNL2d8KUNHMPrQnhCbgjeL1DjsMwbmOpSVZP1Mgp3m2nyZlQHDDMMI6CRY0SJAl+WZkF8SlmoZLUrBUyaptbOMT/tQ03vzFWbNfpC8uwy+SHhVtETV5ElBA0lJacGxtXxLISj+SmLW2EypodJyyZSK7CLyIFlbTX0C3wd7wYtsSlc2EYR1LYIkHdj0bOhGIY5+DzokaXm2OxqGRWr4Wc+Kpie+yBf9By/gzLY/452TbP+3vcGxZRYwwIQNiFs5bHcqNibISKvB+RfGMLfPPzX8iNiYNURvaP5OI0ZrnAXmidQ/g07i5Ea1O4PAzj1RhiDapbaqzhwnwM41h8UtR0GfksKlNMy9UrCLBqaPj32Ddw7DFzg02NyYj43TtsnqcPC7dYVnJjKlm2p97QBBuXrzOLmEqVYQoouYmmKTAIOZXNwsmdCQj5AuM+2YGQWkfxzcRUgEMMGB8gbnk08uMNCDoe4LBjsDuKYRyHT4qa+N3bYV2P0BAUJIJpKX5FJq1uA/w5Y0FBVlAV5FaqDENIaLHjGULDhKvIWwiN+hxPP34jsna8jAXRQ4Hgn109JYZxCkGnAsTiaNgdxTCOwSdFzV8T34K2Zm2Le4hiXor6VfIjonDmjj7wJWR304ujgpG84274IR/h2gxXT4thvBbb7t63cjsFhqkgPilqztxxH/d+KkbQTN38MZJ39sbFdbPEtumRk3BroGMyqjyRk4Z8nDEYUMfPr8xOz4znkV/ZgKyWuQi44IeQQ+a+Y04VNgRbbRimQrg2rYZxK0EzZXQdXFz8ptj2fOhSPB5qttz4OikmEwYkXUXXxCt4KjkJXRKviPXUIp3FGc8mp2GeaJGQcr/zrZOyO8q6nQLDMPbDosbHxYwsaBaNDcGzyR8iD0G4PXAjJkVMd/X03IYRKUnYm5eHTwFQjhvd0vrwlCRXT41xRDo3F95jGI/FJ91PTKF1hqDspjezBuOqqTKa+B0SwcE6DVshZJfTbwWC5omCbXRLDcf75+XhlCGfXVFeAosahvF8WNT4sKARqdoFTIhIQKg2C48Ef4EwbZZL5+dOUAwN0aXI9q4Ft6cNBhY1Xtf3iT8WGcZT4f9eH8xushY0kmRO/CLLzMjwuS6eoftRuyDNf4uVpYbYXHBLQcOMl3XodpGlxjYTChwwzDDlgD+RfdA6803Bts+y+mGX/ia8FTUegRq9i2fontTz80f3wEAMy8sTLqeuBYJmOCC2s5XG+yw1/i50P3EmFMNUDBY1Pupu2pbXEZPSpsMAf9wSuAOPhnzh0jm6M/OiY0VQMMXQyJCgoe2Md2AKNsEUKrlFTE1xhfkIFjcMUzYsanzM3UScNNTFoOT3haDpE7xOxNEwJROl1WJlbJwICqYYGq5T431o9BpUnxIHY4wR2jz3SApldxTD2A+LGh9yNxEppigMSFqJNCkKbf3/xsyoMdykUiEkZFjMeCcaowbBx0vu2eYq2B3FMPbBosZH3E2EXvLHC8lLcdpYFzV1Z7Es5jkEaQpdKgzDuB/sjmIY5bCo8RFBQ0xOm4Yd+g4I02Tgo5hnUEnHxeMYhshqnov8KgYEHwlA4FnHN7QsD+yOYpiycQ/nMeMU7g7+AdHaZCyKHozG/kdcPR2GcRsyO2aLFgnZLXPhzghhE/6+WAaOuZXbKTBMEdhS40N0DtyKPyt35OJ6DFNiNWH3/0hkdxTDlIz7/wczFeJQfhMEQI8G/ifEOgsahrkeQ2xB4b1kz2mRwO4ohrkedj95cTzNZWO8yHTqc+0b7NW3dPX0GMYtkSAVtki45jmihmB3FMPYwqLGSwVNjikIzyR/hMumqqisS0Rdv1OuniLDuCXGCBMkig02AX4pniVqrN1RMixsGF+G3U9e1m2bMEkajEidiwP5LURg8PKYpxGpTXfxTBnGPZGtNLpUrahXwzCM58KixgvTt2dmjMGPuXcjAHlYFv0cavudddkcGcbdMVQqcD0l88chw3g6/F/sZYJmbXZfLMwcKu6/HTUONwfuctEMGcYzCP43ENVfj3P1NBiGUQEWNV7gbpKRJODr7AfF/SFhC/BQiK2vnWGY69HlaBF8zP1aJNgDZ0IxjBkWNV5SLZigHk7LY5/G6uzH0D/kE6fPj2EY1wqbqvgK1Vb15T5RjM/CosYLBE2+5Ad/jbnORqBGjwGhZksOwzBlk9ozU9yG7QqGX6rnZT+VZrXhwnyMr8GixoPdTYRB0mFA8ko08fsPEyNmQKcxuWCWDOO5pNyXAWOMEUEnArxC1BDsjmJ8FRY1HuxuohgaalL5R14X/K1vhydCV6Ee16NhGMVIOgnGKLlFgncIGhl2RzG+CIsaNxUzRGmChliRNRCfZD8FDUxYEDWUBQ3D2Ikh2mguQWoAdOneWYuU3VGML8GixsPcTTKbcrvh9fTXxX1yO/UO3uC0OTKMt2Bpj5Ckg0by3sJ77I5ifAWP+Wly3333oVatWggKCkLVqlXRv39/XLx4Ed7qbipN0BzOb4SXU96DCTo8FvI5BoW+79R5MozXFd7zMtdTaX2iZLidAuONeIyoue222/DFF1/gyJEj+Oqrr3DixAk89NBD8BYxU1b8jEyuFCh6OmVK4bglYDumR04SqdwMw9iPIcacNejP1YQZxivwmP/kkSNHWu7Xrl0b48ePR58+fZCfnw9/f394u7tJJkiTh0kR0zEvYziWxryAAE2+E2bJMN6JIdYzu3MzDOPhosaa5ORkrFq1Ch07dixV0OTl5YlFJj3dfZs6KhE0MncH/4g7gn7m9G2GqSDR6yIQti0EfuksahjGG/AY9xMxbtw4hIaGIjY2FmfPnsW3335b6v4JCQmIjIy0LDVr1oSn8nnWY7hkrGJZZ0HDMBWHxEzw8UD4J3rk77tyx9UMHHOrWOe4GsbbcKmoIReSRqMpdTl8+LBl/zFjxuCff/7Bhg0boNPp8NRTT0GiYi0lMGHCBKSlpVmWc+fOwRP5LudejE2biXuufo8UU5Srp8MwjBcImwXT/hXihoUN40249OfJK6+8goEDB5a6T7169Sz3K1WqJJaGDRuiSZMmwvKyY8cOdOjQodjnBgYGisWT+UffCqNSzKmXfYK/QbRWuZuKYZiSMflLSH4oDX5JfojcGOrVKd3FCRvCun4Np3gz3oBLRU1cXJxYyoPJZHa/WMfMeFuBvQuGang2+UPkIQi3B24U9WgYhlEHQyUDUu/JhCZHg8gNofBFuDAf4214hCN5586d2LVrF2699VZER0eLdO7XXnsN9evXL9FK4+kZT5mmUDydvBxXTZVxo9+/WBg9hONoGMYBhff8qfAefMdKU5KwaTsN2J3BVhvGs/EIURMSEoKvv/4aU6ZMQVZWlii+d8cdd+DVV1/1OPeSkno0RkmLoSkL8J/hRsRpE/FRzDMI1WY7dZ4M4+34UuG9smB3FOMteISoad68OX777Tf4Qj8nIs0UiQvG6ghELj6MeRbV/byrcjLDuFWNmiSP+Bh0CuyOYjwd/m92wwJ7MboUfF3pQRzIb47WAXudMEOG8d1qwmypsYXdUYwnw6LGwShtf0BQurac3RSmzUKHwB1OmCHD+Cb5svspmUVNUdgdxXgqHlV8z1v7OREnDPXQ5coWvJfxEkopvcMwjNodurlFgqJifVX7fc01bRi3hy01buBuIgvNwKQVSJWisTG3J54LW4YAcE8nhnEk1RPikB9rQOAFz+0d59Q4G4KtNoybw5YalbG2zigRNHrJHy8kL8VpY13U1J3FBzHPc5NKhnECfinmFgnaHP4YVOqOkmGLDeOusKXGBdlNMuRmmpCagB36DgjTZGB5zNOopEty8EwZhmEYxjthUeMCd5PMkswX8UXOo9DCiPeiX0Yj/6MOnCXDMDK5dfTI7JCNwJMBCN8Z4urpMAyjEixqKog9wcDW/JffGAkZE8T9qZFTcFvQ7w6aIcMwRclroBctEkL/DmJRwzBeBIsaJ7qbrGnifxjTIyfhhKE+BoaudMAMGYYpiXwuvFehwnwEBwwz7gj/R1c0GLgC4/QP/VTFWTEMoxRDLBfeKw+cCcW4Oxz27yR3E5FtCsarqW+IFG6GYdyhRQKLmvIIG+v6NZwJxbgTbKlxkrvJJGkwInUufsq9C//mNxVtEDS+2xiYYdyj8B6LGpXcUbdixcytbLVhXA6LGie5m97JGCsETQDyMCEigQUNw7gISSOxqFEJdkcx7gaLGge6m2S+zH4IizKHiPvvRI3FzYG7VJwhwzD2YIw0mT/5jIBfKosaR/SJIljcMK7Ap0SNVNBUKScrU9H+YVFrMfnnz/HZ1DRklPOYf+e1xdjUSQDSMSh0CXoFrUWGqZyDMQxTYaQUIPaleBgijcg00GcCN1tTgyP9vsQNy/Lw0OBnxPpn815Cj+oLXT0txkvIzjfafI+XhEYqaw8v4vz586hZs6arp8EwDMMwTDk4d+4catSoUeLjPiVqTCYTLl68iPDwcGg4qAXp6elC5NGbJCIiwtXT8Rj4upUPvm72w9esfPB1877rRlIlIyMD1apVg1ZbcuK2T7mf6EKUpvB8FXrzutsb2BPg61Y++LrZD1+z8sHXzbuuW2RkZJn7cJ0ahmEYhmG8AhY1DMMwDMN4BSxqfJjAwEBMmTJF3DLK4etWPvi62Q9fs/LB1813r5tPBQozDMMwDOO9sKWGYRiGYRivgEUNwzAMwzBeAYsahmEYhmG8AhY1DMMwDMN4BSxqGJw+fRrPPvss6tati+DgYNSvX19EwOv1eldPze2ZPn06OnbsiJCQEERFRbl6Om7LokWLUKdOHQQFBaF9+/b466+/XD0lt2fLli249957RQVVqoD+zTffuHpKbk9CQgJuuukmUTW+cuXK6NOnD44cOeLqabk9ixcvRosWLSxF9zp06ICffvoJngiLGgaHDx8WLSTef/99/Pvvv5gzZw6WLFmCiRMnunpqbg8Jv4cffhgvvfSSq6fitqxZswajRo0SQnnPnj1o2bIlevfujcTERFdPza3JysoS14oEIaOMzZs3Y/DgwdixYwc2btyI/Px89OrVS1xLpmSo0v5bb72F3bt34++//0b37t1x//33i+8DT4NTuplimTlzplDvJ0+edPVUPIIVK1ZgxIgRSE1NdfVU3A6yzNCv54ULzR2bSUBTf5mhQ4di/Pjxrp6eR0CWmnXr1gnLA6Ocq1evCosNiZ0uXbq4ejoeRUxMjPgeICu+J8GWGqZY0tLSxJuaYSpqyaJff7fffrtNDzZa3759u0vnxvjG5xjBn2XKMRqNWL16tbBukRvK0/CphpaMMo4fP44FCxZg1qxZrp4K4+Fcu3ZNfEjGx8fbbKd1cnsyjKMgiyBZTzt16oRmzZq5ejpuz4EDB4SIyc3NRVhYmLAM3njjjfA02FLjxZBpn8zWpS1Fv1guXLiAO+64Q8SJPP/88/BFynPdGIZxLyi25uDBg8LqwJRNo0aNsHfvXuzcuVPECA4YMACHDh2Cp8GWGi/mlVdewcCBA0vdp169epb7Fy9exG233SayeZYuXQpfxd7rxpRMpUqVoNPpcOXKFZvttF6lShWXzYvxboYMGYLvv/9eZJBRECxTNgEBAWjQoIG437ZtW+zatQvz5s0TCSSeBIsaLyYuLk4sSiALDQkaejMvX75cxD34KvZcN6bsD0p6T/3666+WIFdyC9A6ffEwjJpQ3gsFoJPr5PfffxdlKpjyQf+neXl58DRY1DBC0HTr1g21a9cWcTSUMSDDv6ZL5+zZs0hOTha3FDtC5luCfvGQX5qBSOcmU3a7du1w8803Y+7cuSII8emnn3b11NyazMxMEd8mc+rUKfH+oqDXWrVquXRu7uxy+uyzz/Dtt9+KWjWXL18W2yMjI0UNLqZ4JkyYgDvvvFO8rzIyMsQ1JFG4fv16eByU0s34NsuXL6e0/mIXpnQGDBhQ7HXbtGmTq6fmVixYsECqVauWFBAQIN18883Sjh07XD0lt4feQ8W9t+g9xxRPSZ9j9BnHlMwzzzwj1a5dW/x/xsXFST169JA2bNggeSJcp4ZhGIZhGK/AdwMnGIZhGIbxKljUMAzDMAzjFbCoYRiGYRjGK2BRwzAMwzCMV8CihmEYhmEYr4BFDcMwDMMwXgGLGoZhGIZhvAIWNQzDOBTqoyW3SCiOFStWICoqCu5CnTp1RNVje0lKSkLlypVx+vRpOLLrOR3j/PnzDjsGw3gyLGoYhvFJ1BZT06dPx/333y9EkSMbhD711FOYMmWKw47BMJ4MixqGYZgKkp2djQ8//BDPPvusw49FPbNWrVoleo4xDGMLixqG8WLWrl2L5s2bi2Z+sbGxuP3220UzSZlly5ahSZMmCAoKQuPGjfHee+9ZHiM3ikajwerVq9GxY0exT7NmzbB582bLPtTEk77IqRsyHaNRo0aYN29ehedNDQnbtGkjjlmvXj1MnToVBoPB8jjNi+b+wAMPICQkBDfccAO+++47mzFonbbTGNSBfuXKleJ5qampolkfiYO0tDSxjZbXX3/dRqQ888wzoikiNflbunRpqfP98ccfERgYiFtuucVm+7///ot77rkHERERYqzOnTvjxIkTNm65GTNmID4+XliNpk2bJs5zzJgxonFljRo1sHz5cpsxmzZtimrVqolO1AzDFMHVzacYhnEMFy9elPz8/KTZs2dLp06dkvbv3y8tWrRIysjIEI9/+umnUtWqVaWvvvpKOnnypLiNiYmRVqxYIR6n59BHRI0aNaS1a9dKhw4dkp577jkpPDxcunbtmthHr9dLkydPlnbt2iXGoDFDQkKkNWvWWOZBDRjvv//+EudJzQYjIyMt61u2bJEiIiLEPE6cOCEa69WpU0d6/fXXLfvI8/rss8+kY8eOScOGDZPCwsKkpKQk8TjNxd/fXxo9erR0+PBh6fPPP5eqV68unpeSkiLl5eVJc+fOFce5dOmSWOTrQo396DrQtaKxExISJK1WK8YpCTr+HXfcYbPt/PnzYpwHH3xQXJ8jR45IH330kWUcui50LQcPHiy2ffjhh2J+vXv3lqZPny4dPXpUeuONN8R5nDt3zmbsRx99lBtbMkwxsKhhGC9l9+7d4kvy9OnTxT5ev359IQqsoS/RDh062Iiat956y/J4fn6+EBNvv/12icelL+m+ffuWW9RQh+AZM2bY7PPJJ58IASZD83r11Vct65mZmWLbTz/9JNbHjRsnNWvWzGaMSZMmWURNcceVIVHz5JNPWtZNJpNUuXJlafHixSWeA50fdTq2ZsKECVLdunWF8CsOui50LKPRaNnWqFEjqXPnzpZ1g8EghYaGClFmzciRI6Vu3bqVOB+G8VX8ilpuGIbxDlq2bIkePXoI91Pv3r3Rq1cvPPTQQ4iOjhYuKHKDkOvo+eeftzyHXB+RkZE243To0MFy38/PD+3atcN///1n2bZo0SJ89NFHOHv2LHJycqDX69GqVatyz3vfvn3Ytm2bCLy1dnPl5uYKtxC5m4gWLVpYHg8NDRUunsTERLF+5MgR3HTTTTbj3nzzzYrnYD02uaaqVKliGbs46LzJzWXN3r17hbvJ39+/xOeRK0mrLYwCIDcUufhkdDqdcBsWPTa5+uhaMAxjC4sahvFS6Atx48aN+PPPP7FhwwYsWLAAkyZNws6dOy3C4IMPPkD79u2ve55SKN5m9OjRePfdd4X4obiRmTNnimOUl8zMTBFD8+CDD173mLVwKCoWSHyYTCaogb1jU1ZSSkrKdcKjPMdRcmwKEo6LiytzfIbxNThQmGG8GPpC7NSpkxAJ//zzDwICAkSAKVkEKNj05MmTaNCggc1CQb/W7Nixw8aSs3v3bhFcTJBFhYKIX375ZbRu3Vo8Xw6ELS8UIEyWlqLzosXaqlEaFLD8999/22zbtWuXzTpdC7IAqQGd+6FDh66z9vzxxx/Iz8+H2hw8eFAck2EYW1jUMIyXQtYSyqyhL3dyDX399de4evWqRZCQ0ElISMD8+fNx9OhRHDhwQGTazJ4922Ycci+REDp8+DAGDx4sLBKUGURQdhGNv379ejHGa6+9dp14sJfJkyfj448/FvOj7CFydZFF6NVXX1U8xqBBg8R8x40bJ+b1xRdfiLo0stAjqJ4MWYV+/fVXUdSuIu4ccu/RXK2tNUOGDEF6ejoee+wxcY2OHTuGTz75RAi2ikDzJGFJ7kSGYWxhUcMwXgrFmGzZsgV33XUXGjZsKEQBuYnuvPNO8fhzzz0n0qJJyFDcTdeuXcUXf1FLzVtvvSUWitHZunWrSJUmd4ssHshN9Oijjwo3FlXVJatNRSCB8P333wuXGcXFUJr0nDlzULt2bcVj0DlQOjsJObKYLF68WLjeCEq9JsjC9OKLL4q5kyvnnXfeKfec6fqRhYnEkwzFwvz2229CONG1bdu2rXD3lRZjozTdndLMKV6HYRhbNBQtXGQbwzCMqFND4oDcVhUJ/HUXKPB4yZIlOHfunEPG/+GHH0R9GXINKXWTlQcSecOGDcPjjz/usGMwjKfCgcIMw3glVEiQLD1kMaHYHwpgJpeQo7j77ruFi+nChQuoWbOmQ45BbjKyjPXr188h4zOMp8OWGoZhvNJSM3LkSKxZs0ZkCpG7pn///pgwYYJIS2cYxjthUcMwDMMwjFfAgcIMwzAMw3gFLGoYhmEYhvEKWNQwDMMwDOMVsKhhGIZhGMYrYFHDMAzDMIxXwKKGYRiGYRivgEUNwzAMwzBeAYsahmEYhmG8AhY1DMMwDMPAG/g/zNgg9/LCIt4AAAAASUVORK5CYII=",
|
||
"text/plain": [
|
||
"<Figure size 640x480 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"import matplotlib.pyplot as plt\n",
|
||
"import numpy as np\n",
|
||
"\n",
|
||
"from sklearn import datasets\n",
|
||
"from sklearn.inspection import DecisionBoundaryDisplay\n",
|
||
"from sklearn.linear_model import SGDClassifier\n",
|
||
"\n",
|
||
"# import some data to play with\n",
|
||
"iris = datasets.load_iris()\n",
|
||
"\n",
|
||
"# we only take the first two features. We could\n",
|
||
"# avoid this ugly slicing by using a two-dim dataset\n",
|
||
"X = iris.data[:, :2]\n",
|
||
"y = iris.target\n",
|
||
"colors = \"bry\"\n",
|
||
"\n",
|
||
"# shuffle\n",
|
||
"idx = np.arange(X.shape[0])\n",
|
||
"np.random.seed(13)\n",
|
||
"np.random.shuffle(idx)\n",
|
||
"X = X[idx]\n",
|
||
"y = y[idx]\n",
|
||
"\n",
|
||
"# standardize\n",
|
||
"mean = X.mean(axis=0)\n",
|
||
"std = X.std(axis=0)\n",
|
||
"X = (X - mean) / std\n",
|
||
"\n",
|
||
"clf = SGDClassifier(alpha=0.001, max_iter=100).fit(X, y)\n",
|
||
"ax = plt.gca()\n",
|
||
"DecisionBoundaryDisplay.from_estimator(\n",
|
||
" clf,\n",
|
||
" X,\n",
|
||
" cmap=plt.cm.Paired,\n",
|
||
" ax=ax,\n",
|
||
" response_method=\"predict\",\n",
|
||
" xlabel=iris.feature_names[0],\n",
|
||
" ylabel=iris.feature_names[1],\n",
|
||
")\n",
|
||
"plt.axis(\"tight\")\n",
|
||
"\n",
|
||
"# Plot also the training points\n",
|
||
"for i, color in zip(clf.classes_, colors):\n",
|
||
" idx = np.where(y == i)\n",
|
||
" plt.scatter(\n",
|
||
" X[idx, 0],\n",
|
||
" X[idx, 1],\n",
|
||
" c=color,\n",
|
||
" label=iris.target_names[i],\n",
|
||
" edgecolor=\"black\",\n",
|
||
" s=20,\n",
|
||
" )\n",
|
||
"plt.title(\"Decision surface of multi-class SGD\")\n",
|
||
"plt.axis(\"tight\")\n",
|
||
"\n",
|
||
"# Plot the three one-against-all classifiers\n",
|
||
"xmin, xmax = plt.xlim()\n",
|
||
"ymin, ymax = plt.ylim()\n",
|
||
"coef = clf.coef_\n",
|
||
"intercept = clf.intercept_\n",
|
||
"\n",
|
||
"\n",
|
||
"def plot_hyperplane(c, color):\n",
|
||
" def line(x0):\n",
|
||
" return (-(x0 * coef[c, 0]) - intercept[c]) / coef[c, 1]\n",
|
||
"\n",
|
||
" plt.plot([xmin, xmax], [line(xmin), line(xmax)], ls=\"--\", color=color)\n",
|
||
"\n",
|
||
"\n",
|
||
"for i, color in zip(clf.classes_, colors):\n",
|
||
" plot_hyperplane(i, color)\n",
|
||
"plt.legend()\n",
|
||
"plt.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "bac44fa2-7149-42cd-bd4c-b4a1f00512dc",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Использование синтетического дататеста Iris**\n",
|
||
"1) Импорт библиотек"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 67,
|
||
"id": "d9421a11-6724-48cc-a028-8ab37c2607e5",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"import numpy as np\n",
|
||
"import matplotlib.pyplot as plt\n",
|
||
"import pandas as pd\n",
|
||
"from sklearn import datasets\n",
|
||
"from sklearn.linear_model import SGDClassifier\n",
|
||
"from sklearn.preprocessing import StandardScaler\n",
|
||
"from sklearn.model_selection import train_test_split\n",
|
||
"from sklearn.metrics import classification_report, ConfusionMatrixDisplay\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "3c7bd8d3-3af8-44d8-8135-b28a1f35d640",
|
||
"metadata": {},
|
||
"source": [
|
||
"Использование встроенного датасета (Iris)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 68,
|
||
"id": "bcb1d31d-db49-427a-8532-35cf7f835da1",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" 0 1.00 1.00 1.00 14\n",
|
||
" 1 1.00 1.00 1.00 11\n",
|
||
"\n",
|
||
" accuracy 1.00 25\n",
|
||
" macro avg 1.00 1.00 1.00 25\n",
|
||
"weighted avg 1.00 1.00 1.00 25\n",
|
||
"\n"
|
||
]
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 640x480 with 2 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"iris = datasets.load_iris()\n",
|
||
"X, y = iris.data[:, :2], iris.target # используем только два признака для визуализации\n",
|
||
"\n",
|
||
"# Оставим только два класса (бинарная классификация)\n",
|
||
"X = X[y != 2]\n",
|
||
"y = y[y != 2]\n",
|
||
"\n",
|
||
"# Разделим на обучающую и тестовую выборки\n",
|
||
"X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)\n",
|
||
"\n",
|
||
"# Масштабирование\n",
|
||
"scaler = StandardScaler()\n",
|
||
"X_train = scaler.fit_transform(X_train)\n",
|
||
"X_test = scaler.transform(X_test)\n",
|
||
"\n",
|
||
"# Обучение модели\n",
|
||
"clf = SGDClassifier(loss=\"hinge\", alpha=0.01, max_iter=1000, tol=1e-3, random_state=42)\n",
|
||
"clf.fit(X_train, y_train)\n",
|
||
"\n",
|
||
"# Оценка\n",
|
||
"y_pred = clf.predict(X_test)\n",
|
||
"print(classification_report(y_test, y_pred))\n",
|
||
"\n",
|
||
"ConfusionMatrixDisplay.from_estimator(clf, X_test, y_test)\n",
|
||
"plt.title(\"Confusion Matrix on Iris\")\n",
|
||
"plt.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "3f60e638-6260-4216-9694-6535b346af8a",
|
||
"metadata": {},
|
||
"source": [
|
||
"Что показывает данный пример. Мы работаем на датасете Iris я проверил, насколько хорошо обучаемая модель SGDClassifier угадала вид цветка по сравнению с правильными ответами."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "48b728a1-815e-47a4-90d8-a2770e12c40b",
|
||
"metadata": {},
|
||
"source": [
|
||
"Для этого используется отчёт classification_report, который выводит метрики:precision (точность), recall (полнота), f1-score (Среднее между precision и recall — метрика для оценки баланса) support\tСколько примеров каждого класса было в тестовой выборке. Названия цветков хранятся в iris.target_names, Они всё равно участвуют в предсказаниях, просто в classification_report они заменены номерами классов (0, 1, 2). Зачем нужен этот пример? Мы хотим знать не просто \"предсказал вид или нет\", а как часто модель ошибается, на каких классах, насколько серьёзны ошибки.\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "404697fd-01c6-4cd4-98c7-26170472522a",
|
||
"metadata": {},
|
||
"source": [
|
||
" **Визуализация границ принятия решений**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 70,
|
||
"id": "82ff41a1-cb87-48a1-b2a2-47d9c791e583",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/png": "iVBORw0KGgoAAAANSUhEUgAAArEAAAIjCAYAAAAUdENlAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAAWIpJREFUeJzt3Ql4VPX1xvF3sjKTfYeQxSxsKiIQUdmRCIKoqGDrUhHXWrVSbK3YVm3VonUXrYIb1uUv7qhVxIIioCAgKMgiAoYtLAGyk4Rk5v/ca0mJJJCEJHfuzPfzPPOEzA3hACG8OTn3/Bwej8cjAAAAwEYCrC4AAAAAaCpCLAAAAGyHEAsAAADbIcQCAADAdgixAAAAsB1CLAAAAGyHEAsAAADbIcQCAADAdgixAAAAsB1CLADA63z22WdyOBzmSwCoDyEWgO1Nnz7dDDwNPbZu3Wp1iZDMv4sbb7zR6jIA+IggqwsAgJbyt7/9TRkZGYc9Hxsba0k9aL6BAwdq//79CgkJsboUAF6KEAvAZ4wYMUI5OTlWl4FjUFFRYQbXgIAAtWvXzupyAHgxxgkA+N3Yweeff67rrrtOcXFxioyM1OWXX659+/bVeduZM2fq7LPPVnJyskJDQ5WVlaW7775bNTU1dd5u8ODBOvHEEw/7tR588EHz1/rxxx9rnzvuuON0xRVX1Hm7N954w3w749pBxs8xnjPex88Zv5bxax5q/vz5Gjt2rNLS0sxaU1NT9bvf/c7sZDb2z6Shx1133VXn7ZcvX25+sWD8uYWHh2vo0KFatGiRjmXu9bXXXtOf//xndezYUS6XS8XFxfXOxK5fv14XXnih2rdvbwbclJQU/fKXv1RRUVGzfn0A9kYnFoDfMeYyo6OjzYC2bt06PfXUU8rLy6sNTgfDnRHSJk6caL6cO3eu7rjjDjNgPfDAAy1SR3V1tf70pz8d8/sxgnB5ebmuv/56M5h/9dVXmjJlijkLbFxrzihGaWmp+f4O9d1332nAgAFmgL311lsVHBysqVOnmqF63rx5OvXUU5tVv/HFgdF9/f3vf6/Kysp6Rwiqqqo0fPhw8/pNN91kBtlt27bpgw8+UGFhoaKiopr1awOwL0IsAL9jhKQ5c+aYIcyQnp5uhrL3339f5557rvncq6++KqfTWftzfv3rX5uPf/7zn7rnnnvMjuexeuaZZ7R582YNGTJEGzdubPb7uf/+++vUeu211yo7O1u33367+f6NDm1TRzEKCgoOC7FGt/TAgQNasGCBMjMzzeeMLnaXLl3MPz8jyDZ3hGDp0qV1fg8/t3r1am3atMkM5WPGjKl93vjCAoB/YpwAgN8xQt7BAGswwlpQUJA+/PDD2ucODVQlJSVmqDO6kEbHc+3atXXenzFiYFw/9GG83ZEY143up9EVbkzIPJJDay0rKzN//b59+8rj8Zjf/m8Jxu9x9uzZGj16dG2ANXTo0EGXXHKJGWyNLnVzjBs37ogB1nCw0/rxxx8f9c8WgH8gxALwO506darzujEuYISxQ+dXjW+dn3/++WZ4Mr59npCQoMsuu8y89vMZTCPUGtcPfdx5551HrOHhhx82O5BGt/RYGd1WY9bW2MJg/F6MX3/QoEH11tpcu3fvNsOj0XX9uW7dusntdmvLli3Net/1bZSo722M0Y5nn31W8fHx5mjBk08+yTws4McYJwCAnzFmLI0QaIRXo1tq3NRl3Ej09ddf649//KMZ2A5l3JRljAYcyvi297Rp0+p9/0an1JirnTRp0jGv/zI6pGeeeab27t1r1ta1a1eFhYWZ86JGsP15rd7oaF3Ygx566CHz92TcdGd0hX/7299q8uTJ5o1lxk1eAPwLIRaA3zHucjfmUA+9iSk/P18jR440Xzdu8NqzZ4/efvttc1/pQcZMZn2M0Jibm1vnuRUrVjT46xsztREREbr55puP+feycuVKff/993rxxRfN+dSDPvnkE7Uko7trbA4wboT7OaMTbazEMrYitLbu3bubD2M+94svvlC/fv309NNPm3+mAPwL4wQA/I7RITVuUDrI2E5gbAowbm4yBAYGmi+NmdJD7443buo6VsbIgvHrGZsRGtuBPJL6ajV+/Nhjjx3z+/75rzNs2DCzC3ro2MXOnTvNm+D69+9vdq5bizFva/wdHcoIs0Z4NjYWAPA/dGIB+B0jkBr7TS+66CKzs2iEUyOEHdxMYNwUFRMTY95wZHzL2li79dJLL9UJis1l3MFvzJCOHz/+qG9r1DZr1qw6zxldYyO4Gc+fddZZ5viAMe5grKcyRgiMIPnWW28dtve2JRjdTqPDa/xZ/eY3vzFvhjNWbBkh8h//+Idak7HizLgJztiH27lzZzPQGn8nRrg2dscC8D+EWAB+54knntArr7xirmcyOrIXX3yxHn/88dodscauVWP/6C233GJ+29oItMZNXUbwNW4oOlZ///vfazuoR2LcxGQ86mN0jY1QbWxZMFaDHZwPNWZ3jRvSjMDXo0cPtaQTTjjBPFjBmOU1fi1j3tbYDfvyyy83e0dsYxm/F+PP3vi9GmHdGG0wnvvoo4902mmnteqvDcA7OTwt0VoAABswDjAwOqBLliyx9fG0xsyuMdPLp28A/oyZWAAAANgOIRYAbMZYy9USYw0AYGeEWACwmZNOOumwG74AwN8wEwsAAADboRMLAAAA2yHEAgAAwHb8ak+ssdNw+/bt5nGPB/dBAgAAwHsYk64lJSVKTk42D3dpiF+FWCPAtsXZ3gAAADg2W7ZsUUpKSoPX/SrEGh1Yw4uPvS2XM8zqcgAAgI+bumm1nF3zdFrqFqXERGlgIifMHU1JyX6d1PXm2tzWEL8KsQdHCIwAS4gFAACtZcX6HZqVmKeRoxYpJsKpMRnnW12S7Rxt9NOvQiwAAEBbmB9YoE5dS5WdFKchSf2tLscnsZ0AAACgFcRGuJQeFmd1GT6LEAsAANCCpmxcpaheW5TtXKnM8G5Wl+OzGCcAAABoyTnY3INzsMOsLsmnEWIBAACO0VurflBhzhqNTs9TWixzsG2BEAsAAHCs3ddzD3ZfR1ldkt8gxAIAADQD3VdrEWIBAACacfOWK2eThnXJV6+ELG7gsgAhFgAAoIkjBNFJ4RrcJZ+btyzEii0AAIAm2htWaHUJfo9OLAAAQBPnYIen5ynOFW11OX6NEAsAANDILQTDRyxUfGw4c7BegBALAADQ6C0EiWwh8BKEWAAAgCN0YNen7tGlbCHwOtzYBQAAcATO8FDFuVwEWC9DiAUAADjCHOyQTvOsLgX1YJwAAADgZziNy/sRYgEAAA7pvs4PLFD/EbPZQuDlCLEAAACHdF8HJW9UdhJbCLwdIRYAAPi1w7uvnem+2gAhFgAAyN9v3hrZcxHdV5thOwEAAJC/r9DKTuLmLbshxAIAAL+1obJUMXH5VpeBZmCcAAAAyN/nYNPDUq0uCU1EiAUAAH63hcA4SpY5WHsjxAIAAL/rvp7gCtGAVLYQ2BkhFgAA+Dy6r76HEAsAAHzalI2r5MrZZB4hS/fVdxBiAQCAT48QuDOrNaxLvsZkjLK6HLQgVmwBAACfF+dyWV0CWhidWAAA4NNzsKOPXyIp2upy0MIIsQAAwMfnYLOYg/VBhFgAAOB7WwhyF/33KFnmYH0VIRYAAPgEthD4F0IsAACw/QaCWYl5Zvc1JsLJFgI/QYgFAAC2D7Cj+yxRWqwxPsAhBv6CEAsAbcTj8Shv6yaVlBYpMb69khI6WF0SYHsbKkuVnBKjtNhoAqyfIcQCQBv4avlCTZ8xVXnbNtY+d1K3Xrrm0puUmd7J0toAO9vuqlBy+E6ry4AFOOwAAFrZ/MVz9LeHb1Pp9jKdrH46XcN1gk7RpnUb9Ye//UYb89ZbXSJgyy0E9xUt1qDcOeqTuIEurB8ixAJAKzpwoEpPvvCwEh0ddbKnv+IdHRTmiFAHR7p6uwcruDpUz7wyxeoyAdttISjMWWPOwQ7N7KwxGcOsLgkWYJwAAFrRVyu+UElZkU7UqXI4HHWuBTmClObupG/XLNHO3fnMyAJHwRYCHIoQCwCtyAinQQHBCvNE1ns9UrHmy10FOwixwFHGB8zua3oeWwhgIsQCQCuKCI9StfuAKlWhUEe7w67vV2nt2wGov/s6P7BA/UfMVnxsuHolcIQsfsJMLAC0otN691dwUIi2aH29K7e2OH5QescMpadkWFIfYIcVWu70anVJSTRnXwmwOIgQCwCtKCIsUhed+yv9qHVa71mpSk+F+XyZp1irHIu117NLV/zi14fNywL4n5hol9UlwAsxTgAArezi0VeYIXXGzH8pr3qdgh0hOuCpUoQrUrdecZf69OxndYmA124hcOVsUr/kPEnRVpcDL0OIBYBWZgRYI8iec+aF+nLZfPPELuMmrj4n91VwcIjV5QFehy0EaAxCLAC0kfCwCJ05cKTVZQBejS0EaCxCLAAAsBxbCNBUhFgAAGB593V96h6N7LlI2UmJdF/RKIRYAABgeYA1jpAdkNqZ7isajRVbAADAUskpMUqLjSbAokkIsQAAwLI5WKML2yd5ttWlwIYYJwAAABZvIYhmDhZNRogFAABthi0EaCmEWAAA0CbYQoCWRIgFAABtdoSsMT7AFgK0BEIsAABo/e4rR8iihRFiAQBAq6/Qyk7iCFm0LFZsAQCAVhUWvsXqEuCD6MQCAIBW30KQHpZqdUnwMYRYAADQothCgLZAiAUAAC2GLQRoK4RYAADQIuMDsxLz2EKANkOIBQAAx9597U73FW2LEAsAAI5p/tXdrVrDuuTTfUWbYsUWAAA4JjHRLsW5XFaXAT9DiAUAAMe0haBP8myrS4EfYpwAAAAc4xaCLOZg0eYIsQAAoNHYQgBvYZtxgsmTJ+uUU05RRESEEhMTNXr0aK1bt87qsgAA8Kvu67LuSzS6zxJlJ8VpTMYwq0uCH7NNJ3bevHm64YYbzCBbXV2t22+/XcOGDdPq1asVFhZmdXkAAPgsuq/wRrYJsbNmzarz+vTp082O7LJlyzRw4EDL6gIAwB9u3jK6r2mxcRwhC69hmxD7c0VFRebL2NjYBt+msrLSfBxUXFzcJrUBAOBLklNilBYbTYCFV7HNTOyh3G63JkyYoH79+unEE0884hxtVFRU7SM1NbVN6wQAwO62uyoUFr7F6jIA3wixxmzsqlWr9Nprrx3x7SZNmmR2bA8+tmzhHyEAAI2dg72vaLEG5c5Rn8QNdGHhdWw3TnDjjTfqgw8+0Oeff66UlJQjvm1oaKj5AAAATZuDLcxZY+6AZQ4W3so2Idbj8eimm27SO++8o88++0wZGRlWlwQAgE9uIRg+YqHiY8PVK4FDDOC9guw0QvDqq69q5syZ5q7YHTt2mM8bs65Op9Pq8gAAsL35gQXq1LVUXVIS6b7C69lmJvapp54y51oHDx6sDh061D5mzJhhdWkAAPiM2AiX0sPirC4D8K1xAgAA0HqncbmO36RsZ74ktvnA+9kmxAIAgLY4jYujZGEPhFgAAPwUWwhgZ4RYAAD8sPtq3MTVf8RsthDAtgixAAD4Yfd1UPJGZSexhQD2RYgFAMCfbt7K2aRhXfLVK6Ez3VfYGiEWAAA/kpkWr14JQQRY2J5t9sQCAIBjm4MtiytTtnOl1aUALYJOLAAAfrSFIM4VTRcWPoEQCwCAj2ILAXwZIRYAAB/tvq5P3aORPRexhQA+iRALAICPdl/7mN1XthDANxFiAQDwEXRf4U8IsQAA+Ijtrgp16lqqoZl0X+H7WLEFAIAPiY1wWV0C0CboxAIA4CtzsINmK94ZLinV6pKAVkeIBQDAxpiDhb8ixAIAYPMtBCe4QjQglTlY+BdCLAAANkP3FSDEAgBgK1M2rpIrZ5N5hCzdV/gzQiwAADYKsFG9tmhIer7GZIyyuhzAUoRYAABsJDUxWr0SrK4CsB57YgEAsMkcbFlcmbKdK60uBfAKdGIBALDRHGyvhCzmYAFCLAAANthCkLtIMRFO5mCBQxBiAQDwQmwhAI6MEAsAgJcdYjArMY/uK3AUhFgAALxofKAwZ43ZfU2LjeMQA+AICLEAAHiJ7a4KnZgWrwGpQYwPAEfBii0AAADYDp1YAAC8aQ7W6VRm+DCrSwK8HiEWAACrtxB0/2kLAXOwQOMRYgEAsABbCIBjQ4gFAKCNsYUAOHaEWACNcqD6gLblb5bH41FKhzQFB4dYXRJaQU1Ntbbmb1ZNTY06dkhTaEio1SX5XPd1fmCB+o+YrfjYcI6QBY4BIRbAUUPN6++/rPc/flNFpYXmcxFhkRp15oX65XnjFBTEpxFf4Ha7NfPj1/XOh69pT2GB+ZyrXZiGDzlHl114tdqFtrO6RJ9gBNioXlvUJSWR7itwjPjfB8ARg839T96lL5d+rmRPhrLVQw45tLNsq2bMfFEb89brTxPuVWBAoNWl4hj988WH9NHc99RB6eql4xWgQO2u2Kb3Z72p7zes0b23PUL3vYWkJkYrPczqKgD7Y08sgAYtWfGFFi75TCd4+qiro6diHAmKdsSri+NkdfecpsXLF5gBF/a2Zv0qfTR3prqqp05w5CjWkahoR5w6OU7SyZ7+Wv39t/rk8w+tLtMn5mBdx29StnOl1aUAPoEQC6BBH336nqIC4pTkSDnsWoIjWTEBCZo19z1LakPLmfXpewoLiFBHZRx2zfiixfi7/nDOu5bU5itzsPcVLVanc/+tYV3yNSZjGHOwQAtgnABAg7blb1GkO0Zy1H/duLZtx5a2LgstbPuOrYpwx8jhqP8vOtITq227NrR5Xb6ALQRA6yHEAmiQcQPXPsdPN3PVp0L7zbeBvUWERyjfkd/g9QqVK8wV3qY12R1bCIDWxzgBgAYN7nemCpSvck/pYdcqPOXa7dhuvg3sbeDpudrn2a1iz77DrlV5KrUzYAt/z03svhqHGAzKnWNuIWB8AGgdhFgADcodMELxsYlaEbBAezw7zB2xxmOPZ6eWByxQbHSczhx0ttVl4hj1O2WwjkvJ1LcBX2iXZ5vcHrf591zoKTD/7ts5nTrnzDFWl2mbI2TN8YE+SzQ0szPjA0ArYpwAQINczjDd/+cndO+jf9LyvAUKCfhp8X2Vu1IZHbP1p5vvZZzABwQHBeue2x7VfY//Rd9+/6WCA0IUoABVuiuUHJ9irlGLj02wukzbyEyL14DUILqvQCsjxAI4osT49nr07me1Zv1KfbtmueTx6MSuJ+uELj0avBEI9hMTFav7//Kk1m9aq+Url6jGXaOuWSeoxwm9FRDAN+0aOwdbllj23xVaqVaXA/g8QiyAozLC6vGdTzIf8G2dMrqaDzR/C0GcK5ouLNAGCLEAADQTWwgA6xBiAQBoZvd1feoejey5SNlJidzEBbQxQiwAAM3YQuDK2WSODwxI7Uz3FbAAIRYAgKZ2X3MXKSbCqTEZo6wuCfBbhFgAABppu6tCnbqWmjtg6b4C1mJvCgAATRAb4bK6BAB0YgEAaMIWgkGzFe8MZw8s4AUIsQAAHAFbCADvRIgFgHrsryjXZ1/M1jerv5bb7Va3Tt2VO3AEx+z6GbYQAN6LEAsAP/P9xjW684Hfq6S0WNGOeDnk0JdLP9fLbz6r22++R71POtXqEtEG4wOzEvPYQgB4MUIsAByiqKRQf7l/ooIqQtRXZ8mpMPP5Sk+F1hxYprsfmaQnJ7+oju2ZifTp7mv3n7qvabFxjA8AXortBABwiNnzPtD+/eU6yX26nI6fAqwh1NFO3T2nKcAdoA8+ecvSGtG6ATaq1xYN65Kvcd1GEWABL0aIBYBDfPX1QsV5OijE0e6wa4GOQCW6U/Tl0vmW1Ia2kZoYrV4JdNoBb0eIBYBDVFVVKegIk1ZBCtaBAwfatCa03RaCsrgyZTtXWl0KgEYgxALAITplddW+gN1ye9yHXfN4PNoTsFOdMrtYUhtad4ygMGeNRvdZYnZh2UIAeD9CLAAcYuQZo7XfXaaNWm2G1kNt1UYVu/dq1LALLasPLb+F4L6ixRqUO6d2DpYAC9gD2wkA4BCZ6Z105S9/o+df+6cKA3abM7ABCtDugO3a49mp84aPVe/urNjylfEBs/vKFgLAlgixAPAzF559idJTMvT2hzO0cs3XZke2c8bxuvqs32jAqWfI4XBYXSJaoANrnMJ1aZd89UrIovsK2BAhFgDqkdPjdPNhBFjjERDA9JWvcYaHKs7lIsACNkWIBYAjMLqudF59b4zA6MKO7LRIUpzV5QBoJkIsAMC/TuPK+ek0rgGpnenCAjZGiAUA+MUM7KzEPI3MXaSYCKfGZIyyuiQAx4gQCwDwaWwhAHwTIRYA4LPd1/mBBeo/YrbiY8PZQgD4GEIsAMBnxweME7jSYhPpvgI+iJ0xAACfXaGVFhtNgAV8FCEWAOBzjDGCmLh8q8sA0IoYJ4BX2lWwQ/srypUQ114up8vqcmyprLxUu/fsUpgrTAlxSVaXA1iyhWBI0jCrSwLQSgix8CqLv16gV95+XhvyvjdfDw4K0ZB+Z+rysdcpJirW6vJsoWDvLk2fMVXzF89RdU21+VyXzON12Zir1at7H6vLA1oNWwgA/0KIhdeYPe8DPfbsfYp1JKq7TlOo2mlf9W7Nmz9H3363XA/dNVXRUTFWl+n1AfZ3d1yr8pJyHefupmjFqUL7tW3TRt3xwC269Td3aeBpQ60uE2hRbCEA/BMhFl6hpKxY/5z+sJJ1nLp5etce8xmteLV3p2np3k/1yjvP6YYrfm91qV5t+oynzQCb4x6idg5n7fNJnhStcnylKc/9Q3169lO70HaW1gm0dPd1UPJGZSexhQDwJ9zYBa/w2RefqKamWlk68bBz6p2OMCW7MzRn/ixVVlVaVqO3Ky0r0fzFc5XqzqoTYA3Gn2mW5wSVV5Rp4ZJPLasRaMnuq3GEbOaI9zSsS76GZnYmwAJ+hhALr7B9xxaFB0Qq1FF/h9DoyFZWVaiwaG+b12YXu/fsNGdgjT+r+rgc4XIFhmv7jq1tXhvQGtzp1eqSkqgxGcMYHwD8ECEWXsHlClelp0Juj7ve6xUqN18627GpoCFhrnDzpTEDW58aT7Wq3BVyOcPauDIAAFoeIRZeYUCfIap0V2iHNh92zQi22wM2qUe33oqMiLKkPjtIjG+vTsd11VbHBnk8nsOub9ePqvHUqN8pgy2pD2jJOdhl3Zdo+PELrC4FgIUIsfAKx6VmmeFqnWOFtnk2mWHLUO4p1SrHYpV4CnXx+VdYXabXu/TCq1ToKdB3WqL9nrLaDuwWzw9a71ip3IEj1D4x2eoygRaZgx2QmsUcLODHHJ76WjY+qri4WFFRUXpj2sd8S9ULGTdtPTrt7/p88RwFB4Qo2BGi8ppShTnDNeHaSeqbM8jqEm3h04Wz9eQLD6iiskLOwDBVeSpV4642A+wN4/+g4KBgq0sEmtV9XZ+6RyN7LlJ2EjtgAV9WUlyujI7XqqioSJGRkb4RYj///HM98MADWrZsmfLz8/XOO+9o9OjRjf75hFh72Ja/WQuXzNP+ynKldkhXvz5DFBoSanVZtmKcdrbgq0+1fedW84uAfn0Gq0NiR6vLApq/A3bQbDldIWb3lZu4AN/W2BBrqz2xZWVl6tGjh6688kpdcMEFVpeDVtKxQ5ouOvdXVpdha8YNcGcOPNvqMoAW7L6yAxaAjUPsiBEjzAcAwPc7sEaAHd1niQakdqb7CsDeIbapKisrzceh4wQAAHtwhocqLTaaAAvA/7YTTJ482ZyBPfhITU21uiQAQCPnYId0mmd1KQC8mE+H2EmTJplDwQcfW7ZssbokAMBR5mBnJeZpUO4cthAA8N9xgtDQUPMBAPB+xg5YV84mjU7PYw4WgH+HWACAjbYQ5C5STIRTYzJGWV0SABuwVYgtLS3VDz/8UPv6pk2btGLFCsXGxiotLc3S2gAATUf3FYBfhNilS5dqyJAhta9PnDjRfDlu3DhNnz7dwsoAAE29ecuYfaX7CsAvQuzgwYNlowPGAJ9RUbFf8xb9R9+uWS55PDqxaw8N7jvMPFQBOJYVWty8BcAvQiyAtrf2h+/014duVXFpkaID4uWQNO/L/2j6jKm6Y+J9OqFLD6tLhA1tqCy1ugQANufTK7YAHJu9hXt0xz9uUWB5sPpphHI8g9XbM9j8cch+p+584Pcq2LvL6jJhwznYwpw1Gn78AqWHxVldDgCbIsQCaNCsT99TZUWlTnL3ldMRVvt8O4dLJ3lOV/WBGn04511La4S95mDvK1ps7oAd1iVf47qN4kYuAM1GiAXQoC+Xfq54TwcFO0IOuxbkCFaCO1lfLvncktpgv+7rsu5LNLrPEg3N7KwxGcOsLgmAzTETC6BBVVVVCtbhAfYg41pFVUmb1gR7YQsBgNZCiAXQoOyMzlqya7E8bo8cDuOWrv8xNoXsDdil7hnc2IWGDzEwZl+NHbBpsWwhANCyGCcA0KCzc89XmbtYefr+sGtb9INK3IUalXuBJbXBHqdwGbOvA1KzCLAAWhydWAANOr7zSbro3F/p9fde0j7HLiW4O8ohh3Y5tmqPdur8Eb9UjxN6W10mvFRySoziXC5u3gLQKgixAI5o3NjrlJXeWe98OENrN3xtPtf5uG66asT1GnjaUKvLg5fOwRpd2NHJSyRFW10OAB9FiAVwVP37DDEf1dXV5utBQXzqQMNbCFzdN/13DjaaMQIArYb/iQA0GuEVDWELAYC2xv9IAIBjwhYCAFYgxAIAmt19nR9YoP4jZis+Nly9ErK4iQtAmyHEAgCavUJrZM9Fyk5KpPsKoM2xJxYA0OwVWtlJjA8AsAYhFgDQLGHhW6wuAYAfY5wAANDkLQTDRyw052DTw1KtLgmAnyLEAgCasYWAOVgA1iLEAj5qf0W59u4rkMsVrpio2GN+f263W7sKdsgjjxLj2yswIFC+zvg9796zU26PW4lxSQoM9M9PmWwhAOCNmvQZ+cMPP9Tbb7+t2NhYXXnlleratWvttX379unCCy/U3LlzW6NOAI20t7BAL735jD5d+IkOVFeZz53Y5WRdcsGV6nF8r2YFuZkfv6GZs17X7r07zefiohN03lljNXrEL3wyzHo8Hv37P2/r7Q9f086CfPO5mKg4nTPsQl048hK/OvSBLQQAvJXDY3y2boRXX31Vl19+uc466ywVFRVp6dKlevbZZ3XppZea13fu3Knk5GTV1NTIWxUXFysqKkpvTPtYLmeY1eUALW7PvgJNvPNalRSVKMWdqSjFqVL7tS1go4o9+3TbTXer3ymDGv3+jE8PD0+7V3MXzFIHpStJKcanDe3UVu1QngacNlR/uP4OBQT4zj2ixu/5n9Mf0odz31V7pam9UuWQQ7u0TfmOPPXp2Ve333yvT4b3eo+QPX6T+qXnaUAq3VcAbaOkuFwZHa8182ZkZGSDb9fodsIDDzyghx9+WL/97W/N119//XWzG1tRUaGrrrqqZaoGcEymz3hapUUlOsU9RO0crtrn27vTtNKxWI89M1k5PU5TaEhoo97fspWLzQB7gk5RB0d67fPxaq94T5I+X/QfDTo9V6f18p3u3HfrvjEDbFf1Uoojs/b5OPP33EGLvl6gBYs/NX/fvt6BdXer1rAu+RwhC8ArNbp9sn79ep1zzjm1r1900UV6//33NWHCBD399NOtVR+ARiorL9Xni+YoxZ1VJ8AaHA6Hsj0nqmx/qRYu+azR7/OjOTMVGRBjdiR/LsmRquiAOH005135ko/mzlR4QKQ6KuOwawmOZMUGJOrD//jW77khMdEuxbnqfiwBgLdodCfWaOcaIwMZGf/7xD5kyBB98MEHGjVqlLZu3dpaNQJoBOMGpOqaA4pRQr3XXY5wuQLCtS1/c6Pf55ZteYpyx5khuD7Gtc3bfpQvOdrvOdodry35efKHLQT9ko3fZ7TV5QDAsYXYPn366KOPPtJpp51W5/lBgwaZHVkjyAKwjrPdTx2zCu1XVD3XazzVqvJUNmke3OV0qUilDV6vVIXCnOHyJS5XmIocOxq8bswY++pMPVsIAPjkOMHvfvc7tWvXrt5rgwcPNoOsceMXAGskJXRQVlpnbXNsNG9O+rl85anGXa2+OY2/sWtQ31wVOPJV4Sk/7FqlZ792O7ZrYN+h8iUDTxuqPZ4dKvccHt6NLwJ2BWwz/1x8sftqHGIwKHeOuqQkakzGMAIsAN8IsUbHddKkSQ1eN0YLXnjhhZaqC0AzXHrhldrr2aXVWmaGTEONp0bbPJu03rFSQ/oNU4ekjo1+f7kDRyo6MkYrAhaq0FNghmPjUeTZYz4XER6p4YP/NyvvC4w/o4S4JH0TsFD7PLtrvyAwtjsYzxlfzJ89dLR8bQuBeYhBnyUamtmZNVoAfGvFli9gxRb8wSef/1tPTX9YVQeqFBYYaYbZA+4qDT59mG6++o8KaeRmgoOMGdq/Pnybtu3YLGdgmLluqrymVB0SO+qOifcprePhN0DZ3Y5d2/W3h29T3raNdX7PibHt9ZeJk5WZ3km+dISssQM2JsJpdl8BwC4rtgixgM9uKviPtu3YqjBXuAaceoZSOhy+YaApBx4sX7VE36752uxMdu/aU71POtWn9sP+nPH7/Gb1Mq1YtdQ8sev4zifplJNP95n9sAcPMTC6r2mx0XRfAfjunlgA9mEE1xFntNy3vI2waoRW4+EvjO0EJ5+QYz58VXJKDAEWgG35bhsFANDgGMF2V4X6JM+2uhQA8I4Qu3r16pZ8dwCAVtxCQBcWgF+F2CuuuMKcjzuU8fq9996rU045pSVrAwC0ILYQAPAlTZ6JXb58ucaOHavXXntNwcHB+u6778xgW1hYaB6GAADw0i0EuQe3EHA4DQA/7MR+9tlnys/P18iRI3XPPfcoJydHp59+ur755hsNHDiwdaoEADR7fGBZ9yVm9zU7KY41WgD8txMbExOjTz75ROedd57uvPNOvfXWWxo92rcWfwOAz3Rfz6X7CsA3BTRn12pNTY1effVVnXHGGWaQzcvLM583HgAA622oLDVXaNF9BeCrmtyJjY6ONvcnGg6ek5CZmWn+2HjeCLgAAACAV4XYTz/9tHUqAQC02BYCV84m9UvOU3pYltXlAIB3hNhBgwa1TiUAgGPCFgIA/qTZx86Wl5dr8+bNqqqqqvP8SSed1BJ1AQCauIXA3AGbnqe02Dh2wALweU0Osbt379b48eMb3AnLTCwAWDM+MKxLvnolZCkzvJvVJQGA920nmDBhgnmwweLFi+V0OjVr1iy9+OKL6tSpk957773WqRIAcESZafHqlZBKgAXgN5rciZ07d65mzpxpHnIQEBCg9PR0nXnmmYqMjNTkyZN19tlnt06lAIB6xwjKUsuU7VwpKdXqcgDAezuxZWVlSkxMrD34wBgvMHTv3l1ff/11y1cIAGhwjMCcg+2zhC4sAL/T5E5sly5dtG7dOh133HHq0aOHpk6dav746aefVocOHVqnSgBALbYQAEAzQuzNN9+s/Px888fGaV1nnXWWXnnlFYWEhGj69OmtUSMA4L/YQgAAzQyxl112We2Pe/fubR45u3btWqWlpSk+Pr6p7w4A0Mju6/zAAvUfMVvxseFsIQDg95q9J/Ygl8ulXr16tUw1AIAGu6+DkjcqOymR7isANCfETpw48YjXH3744WOpBwDwM9tdFToxLV5D04LovgJAc0Ps8uXLa3+8YMECc6TA2BdrcDgcTX13AAAAQOuH2E8//bT2xxEREXr11VeVmZnZ9F8ZAND4LQROpzLDh1ldEgD4zkwsAKDlsYUAAI6MEAsAXoQtBADQSiH2vffeq/2x2+3WnDlztGrVqtrnzj333Ka+SwDAf7uv61P3aGTPRWwhAICWDrGjR4+u8/p1111X+2Pjxq6ampqmvksA8GuHdl9PcIVoQGpnuq8A0NIh1ui+AgBajhFgo3ptUZcUuq8A0FjMxPqhisoKFZcUKswVbj4AWC81MVrpYVZXYQ8HDlRr545CtWsXoviESKvLAWCXEFtQUKDbbrvNHBt46KGH9Mwzz+iVV14xT+16/PHHFRnJJxRvtatgh155+3nN+/ITHag+YI5/9Dm5ny65YLyyj+tidXmA387BunI2KduZb0RZq8vxaqWlFXr4H+/q5Rfnau/ecvO5Pqdm6eZbRmv4iJ5WlwegjTk8Ho+nKT/hoosu0qZNmxQXF6eKigrt2bNH11xzjZ5++mkNGDBAU6dOlbcqLi5WVFSU3pj2sVxO/2p55O/apt/fdb0qyirU0Z2hSMWoXGXaFrBRlQH7dfetD6l7N/4TANp8DnbQwS0EqczBHiXAXjDqXq1bs1lXXxqhYYNd2rOvRs++UqLPvyzXA4+O1/irhlpdJoAWUFJcroyO16qoqOiIzdEmd2Lnzp2r2bNnKzs7WzExMfrkk090xhln6IQTTtAVV1xxrHWjlUx96TFVlVXpFPcZCnW0M5+Lk5TsPk4rPAv08NR79dzDrysgIMDqUgGfxxaCpvvnlA+1ZnWe5r3TUb17/PQ5zHDJBRG6cdJu3f6HFzVyVG8lJUVbWieAttPkxFJWVqbExEQzGbtcLqWnp5vPd+7c2Rw1gPcp2LtLS1d8qXR3l9oAe1CgI1DZnhO1a88OrfhuqWU1Av5iysZVPx1i0GeJhmZ2JsA28obifz3/H10+NqJOgDUYY1H33BanoCDptZc/t6xGAG2vyZ3Yjh07Ki8vTykpKfroo4/Ml4adO3ea4RbeZ9uOrfLIoxgl1Hs9UrFmmN2y/Uf16t6nzesD/Kr7mmt0X40TuEZZXZJtlJZUaMeOYg3q277e6zHRgep5Yjt9v257m9cGwEYhdvLkyeZcqaF///91EDZs2KDx48e3bHVoEc52TvNllSrk0uHbCKp1QDWeGrULdVlQHeAf3Vfj5i3jCFl2wDZdaLtgBQQ4tHNXdb3XjVs7du6uUdeT63ZpAfi2JofYsWPH1vv8L37xi5aoB60g67jOio9J1NZ9GxSt+MOub9VGBQYG6dSe/SypD/AHmWnxGpAaRIBthtDQYHP7wLOvrNb1V0QrONhR5/rsz8q14cdKPTia7yQB/iSgOXf4H+kB7xMYEGiu0dqhLVrv+VYHPFXm80b3dYvnB21yrNbIM0YrOirG6lIBn9xCYMh2rrS6FFubcMu5+n5jlX553Q7lbTlgPud2e/T+7FL96sZdOq1vJ/UfwBcIgD9pcic2OjraHKSv79s5HDvrvYYPPkfFpUV66Y1ntNWzUWEBEarwlKvKXalhA0fp6ktutLpEwKe3EMS54ujCHoPep2TruX/drBuve0rZp/2oLtntVFjkVv7OKvUf2FUvvDSh3v+bAPiuJofYzMxM7dq1yzzwoF8/vv1sJ2NHXabcASM1d8Es7d6zS5HhkRrU90x1bM+CdaClMQfb8owVWt+ue0LvvPml1qzeYp7YNXJUjnqfkkWABfxQkw87OHDggKZMmaJ7773X3A/7j3/8QxkZGbIDfz7sAEDbd19jIpwakzHM6pIAwCcPO2jyTGxwcLAmTpyo9evXm+u2TjrpJN1yyy0qLCw81poBwKd2wBJgAaD1NPt4ptjYWD366KNavny5fvzxR/MEL+N1APDHm7fuK1qsQblzNKxLvsZ1G8X4AAB420xsz549D5s9MiYSKisrzY7shAkTWrI+APB68wML1Klr6X8PMeAELgDwyhA7evTo1qkEAGwsNsKl9LA4q8sAAL/R5BB75513tk4lAGDXLQTHb1K2M18Smz4AwGtD7EFLly7VmjVrzB8ff/zx6t27d0vWBQBePwc7KzFPI3PZQgAAtgixW7du1cUXX6yFCxeaBx8YjM0Effv21WuvvaaUlJTWqBMAvKv72v2nHbBpsczBAoAtthNcffXV5q5Yowu7d+9e82H82O12m9cAwJ+2EBBgAcAmndh58+bpiy++UJcuXWqfM35sHIAwYMCAlq4P8DpVVZVasOQzrf7+Wznk0EnH99RpvQcqOCjY6tLQyocYmDtg6b4CgD1DbGpqqtmJ/bmamholJye3VF2AV/p+4xr99cE/qrBkryIDY+SRRx/OfVfxMYn66x8e0HGpWVaXiFY8QtbovvZKyGIHLADYcZzggQce0E033WTe2HWQ8eObb75ZDz74YEvXB3iNPfsK9Of7fieVOXS6hqmPe6hOdefqNJ2pqqIDun3yBBWXFFldJlpJZlq8eiWkEmABwK4h9oorrtCKFSt06qmnKjQ01HwYP/7666915ZVXmid5HXwAvuTDOe+oqrJSPdz9FOb431nO4Y4o9XD3VWlpsT75/N+W1ojWmYMtiytTtnOl1aUAAI5lnICjZeGvFiz+TAnujgp2hBx2LdThVLyngxZ+9ZkuPPsSS+pD687Bxrmi6cICgJ1D7Lhx42SlJ5980hxp2LFjh3r06GHeUNanTx9La4J/qKjcr3DFNHg9WKGqqKho05rQujtgh49YqPjYcOZgAcDOIba4uLhRbxcZ+b9vs7a0GTNmaOLEiXr66afNEQajKzx8+HCtW7dOiYmJrfbrAoaMtCx9X7RO8hx+zePxqDBgt3qn8wWVb20hSGQLAQDYPcQaBxs4HI4Grxv/iRvXjS0FreXhhx/WNddco/Hjx5uvG2H23//+t55//nnddtttrfbrAoaRuedryTe3art+VLLjuDrXtmiDSt3FGjn0fMvqw7F3X+cHFqj/iNl0XwHA18YJ3nzzTctu2KqqqtKyZcs0adKk2ucCAgKUm5urL7/8st6fU1lZaT6a2k0G6nNKj9M1bNAozZ73gQqUr0SPcTqdRzsdW7Vb2zX6rF/oxC49rC4Tzey+rk/do5E9Fyk7ie4rAPhciO3Xr59l37YvKCgwu7xJSUl1njdeX7t2bb0/Z/LkyfrrX//aRhXC1xnfabjpylvVObOr3v3oda3asdh8Pj05U787+08a2v8sq0tEM213VahT11INzexM9xUAfPXGLjsxurbGDO2hnVjjsAaguYzu/4gzRuusIeeptKxEckjhrogjjtrAHmIjXMb3fKwuAwDgayE2Pj5egYGB2rlzZ53njdfbt29f7885uMcWaGlGaI0Ib72bGNHGc7CDZiveGW6cSWh1SQCAlj7swPhP28puU0hIiHr37q05c+bUPud2u83XTz/9dMvqAmDfOdhl3ZfozBGfqUtKosZkDGOUAAB8sRNrbB8wTus6Wmfz7bffVmsxRgOMPbU5OTnmblhjxVZZWVnttgIAOBq2EACAn4VYqw85MPziF7/Q7t27dccdd5iHHZx88smaNWvWYTd7AUB92EIAAL7D4TFarH7CuLErKipKb0z7WC5nmNXlAGhDUzaukuv4TeqXnqcBqXRfAcBblRSXK6PjtSoqKjriIVq2ubELAI4lwEb12qIh6fkakzHK6nIAAC2AEAvAL6QmRqtXgtVVAADafDsBANh1DtYYI8h2rrS6FABAC6ITCzSDsd5tW/5mKcChjkmp5iEI8PYtBKk+OQdbXV2jPXtK5HKGKCLSOLABQFuprDygfftKFRnpksvFXvq2RogFmqC6uloPPX23vljyuardB8znggNDNKjvUN189STCrJfwhy0EJSX79dhD7+ml6XO1Z0+Z+dygIcdrwu9Ha8DA460uD/BpW7cU6MH739Vbry/U/v0HFBwcoFHn9tHvbztfXbp2tLo8v0GIBZrQff31Hy9T/q6tSlRHJSlVHnm0o2az/jP/I23IW6/H736eIOsNWwhyNmm0uYWgs092X40AO/rse/TDuq266pIIDenfQbsKavTMy5t04TmT9c9nrteYi/paXSbgk37ctEtnn3mnHJ4K3XZjhHqcGKr1Gw/oqenLddYZy/XW+7erV+8sq8v0C4RYoJFefed5M8CeoFPUwZFe+3x7pWqL5wet27xC789+U+eddZGldfrz+MCsxDyNzF2kmAinT28hePTB98wA+/nMjjr5xP99C3P8LyM1/uadmnjTMzpzWA9FRbNKEGhpf5z4gsKdVVrwfooS4/8Xo679VZRyx27Xzb+Zqs8X3W/pKaf+gpYR0EgfznlXEYquE2APSlGWnArTO7NmWFKbvzO6r8YRsqP7LFF2Upx5hKyvMmZgX35xrtmBPTTAGgIDHbr/L/GqqqrWGzMWWlYj4KvyftylOf9ZqdsnRNcJsIbwsADdOylWa1Zv1+Ivv7esRn9CiAUaqbS0VHGq/3Q44yvuOLVXUdG+Nq8LP8lMizcPMfDF+ddDFRQUmzOwQwc6673eISlIJ3Rtp7VrtrV5bYCvW/99vvkyd0D9//6G9HPKaMCuXbu1jSvzT4RYoJGMWdcqVTZ4vUoVCgwMbNOa8NMYgcFfVmi5nD91X3furqn3utvtUcGeGu6UBlrBwX9XDf37K9hbI+McVP79tQ1CLNBInbO7aYe2qMpzeJCt9OzXbm3Xid1OtqQ2f95CYMzBDsqdoziXyydv4vq5yCiXBgzqpmdeLjED68/9+z9l2pZfpVHn5lhSH+DLcvpkKzExQtNeKqr3+rR/FalduyDlntmjzWvzR4RYoJF+/aubzW0EyzVfpZ7i2udLPIX6WvPNkYLrLptgaY3+NgdbmLPGnIMdmtnZ58cIDvW735+npSv2mzdx7dxdbT5nBNr3Z5fqygm7zZB7yqmdrC4T8DkhIUH67cTz9Owrxbrn4T0qLXObz1dWuvXU9ELd/cg+jb/6TMXGRVhdql9weDxG49s/FBcXKyoqSm9M+1guJ3ftounmLpylR6ZOlttTI5fCzVC7X2UKdARp0s1/0+m9B1pdov9sIeh5cAuB797EdSTGjVsTb3rW3F18Yrd22l1Qo63bq9R/YFdNf/l3io7hcxzQGozYNPnuN/XoQ+8pzBWg7MwQ5W05oL37qnXZuMF68NHxCgpitOxYlBSXK6PjtSoqKlJkZGSDb0eIBZqovKJc/3p9qr5d/bVxR5d6n3SqLj3/KrVr187q0vwmwBrd17TYaL/qvtancF+ZXn9tgdat3WbO4J1z3ilmB5bVPkDbHHgw4/8WaNvWPUpIiNSYi/qpU5dkq8vyCYTYehBiAfuH2HmZO3RN/1V+H2ABwN9DLDOxAGxjfmCBYuJ+WnEDAPBvnNgFwBZbCNan7qk9jWtIkn/OwQIA/ocQC8DrtxC4cjZpdHqeBqR29os1WgCAoyPEAvDuLQT/7b6OyRhldUkAAC9CiAXgleMD5g7Y9DylxcZxExcA4DCEWABe1X01bt7qP2K24mPD1Sshi/EBAEC9CLEAvIo7vVpdUhLpvgIAjogVWwAAALAdOrEAvGcLQfdNGp6ep/SwLKvLAQB4OUIsAEuxhQAA0ByEWACWYQsBAKC5CLEA2hxbCAAAx4oQC8CaI2R7LlJ2ElsIAADNQ4gF0KYdWCPAXjrwG/VK4AhZAEDzsWILQJtyhocqzuUiwAIAjgkhFkCbbiEY0mme1aUAAHwA4wQAWh1bCAAALY0QC6DVsIUAANBaCLEAWgVbCAAArYkQC6DVuq8nuEI0IJUtBACAlkeIBdDyR8jSfQUAtDK2EwBo8RVa2UncvAUAaF2EWAAtZkNlqWLi8q0uAwDgBxgnANDiWwjSw1KtLgkA4OMIsQCOCVsIAABWIMQCaLYpG1fJlbPJPMSALQQAgLZEiAXQ/O5r7iLFRDg1JmOU1SUBAPwMIRZAk9B99X0lJfvlrnErMsolh8NhdTm2dOBAtUqK9ysi0qngYP6rReN4PB4VF5UrIDBAERFOq8vxevzLAtCkG7jcmdUa1iWf7qsP+mDmEv1zygf6avEG8/XsTkm6+rrhGn91rgIDWWbTGD9u2qVHHpypt9/4Qvv3H1BYWIjG/KK/fvf7c5WSGm91efBSNTVuvfDsf/Ts1I/1w/qd5nN9Ts3Sb24apVHnnWJ1eV6Lz0oAmizO5bK6BLSwRx6YqSsue0xhwfl67tEkvfJUe/U6vly33/qSfn3Vk3K73VaX6PXWrN6iYYP/rM8++VKTborUG8920C2/Dtes9+dr2OC/aMMPO6wuEV4aYK+98gnz31rvE8rNf3vGv0Hj36Lxb/LRB9+zukSv5fAYvWs/UVxcrKioKL0x7WO5nGFWlwPYeAsBhxn4ktXfbdHA0ybpLxNjddcf4upce+uDEl10zQ49/dxvNOaivpbVaAfDh/xFlWX5+uztZMVEB9Y+v6ugWoNGb1dSSqbe+eBPltYI7/PGjIW6/uqnzC96Ljg7vPZ5I57d9cBe3fPIXi346j517ZYif1FSXK6MjteqqKhIkZGRDb4dnVgAjZqDLcxZo9F9lmhoZmcCrI/51wtz1T4xWH+aEHvYtQtHRWhIf5emP/uJJbXZxbff/KhlSzfp7j/G1AmwhsT4IP1lYrTmz1ujH9ZzGAjqMv5tDR0YVifAGox5dOPfZFJCsF58fq5l9XkzQiyAI87A3le0WINy55hzsOO6jeJGLh+0+rvNGtwvVMHB9d/ENWyQ0/xWORq25ruf/nyGDap/1GbYoJ+++7d29dY2rQs2GUMZVP9NXCEhDvPf5prVm9u8Ljvgxi4ADW8h6P7TFoK0WMYHfJnLGaqCPQ3PvBbsrZHLFdKmNdmN0xVa+2eVknx4f8h4/qe3488RdTmdIbUfH/Up2OuW0/nTxxfqohML4KjdVwKsbxsxKkdzF5Rrw49Vh10rL3frlbfKNGJUH0tqs4vBQ06UyxWsZ14urvf6My8XKTraqdP7dW3z2uDdRp7TRy+/War9+w//QvKHTVWaO79cI0flWFKbtyPEAqgTYGcl5pmzr8bNW2MyhlldEtrAhRf1VceUWJ13+U6tWFVZ+3zelgO64ModKi6Vrr1+uKU1ejtjp+5V1w7X5Mf36ekXC1VV9dM90xUVbj02bZ8ee6ZQ190wUq7/dmyBg4x/W0UlMv+tbd56oPb55SsrdN64HUpJjdUFY0+3tEZvxTgBgDqc4aFKi42m++pHwsPb6Y13J+mSMf9Q7zM364Qu7dSunUPLV+5XZKRTL7/+e2V36mB1mV7vT3eO1b59pbrhts901wOFysoI1vqNVdqzt1rXXHembrn1PKtLhBfq1DlZL8+4RVf+6jFlnfqjenZ3qqLCo+/WVSgrK1FvzrxVYWHtrC7TK7FiC0CdOdioXls0ptseQqwfqq6u0ccffq25c1aaP+6dk60LLzqd/0CbaN3abXr9/xZo185CdUiO1UUX9+eLABxVaWmF3n7jSy1b+oOCggJ1xtDuGj6yl/ljf1PSyBVbhFgAtWMExg7YmAgnYwQAAK8PsYwTAH7OOMTA3AHLFgIAgI0QYgF/776ee7D7OsrqkgAAaDRCLOCH6L4CAOyOEAv4Wfd1fmCB+o+YrfjYcPVKyOIELgCALRFiAT+yobJU7m7V6pKSSPcVAGBrHHYA+JmY6PrPdgcAwE7oxAJ+NgfbLzlPUrTV5QAAcEwIsYCfbCEYPmIhc7AAAJ9BiAX8ZgsBc7DAzxnn/XyxcK3mfvKtqqqqdXKvDI069xSFhgZbXRqAoyDEAj6ILQTA0eXn79O4ix/W18s2Kbl9iFyuAD31xEdKTIzQsy/erL79u1pdIoAj4MYuwAe7r8b4wKDcOeYWAuMIWQIsUJfRdf3F+fdp5/ZtmvVasjZ/naZ1C1P13efpOr5TjS4e8w+t/3671WUCOAJCLOBjAXZ96h6N7rNEQzM7Mz4ANODDD5Zq9Xfb9M4LSTpzUJgcDof5fNdOIXrvXx0UFeHR00/OsrpMAEdAiAV8THJKjNJio+m+Akfw/swlOqWnU717tDvsWpgrQOMuCtd77yyypDYAjUOIBXxoDna7q0J9kmdbXQrg9cpK9is5qeH/AjskBam0tLJNawLQNNzYBfjQFoJByRuVFhvHGAFwFFmdk/XOjDWqrHQrNPTwMPvZF/vVqXN7S2oD0Dh0YgGbd1+nbFylzBHvaViXfOZggUYad8UQ7So4oH88ue+wa/O+KNfMWWUad2WuJbUBaBw6sYDNb+Ia2XORspPYAQs0ReeuHXXr7Rforr+/raXfVOnyseEKDwvQB5+U6fn/K1G/AV112bjBVpcJ4AgIsYANGd1XV84m8xCDAamduYkLaIZbJ12g4zIS9eRjH+iia7aazxk7Ym+ccJ4m3HIOBx4AXo4QC9jwCNmRuYsUE+HUmIxRVpcE2NpFv+yvsb/opx35+8zdsckdYxUczH+NgB3wLxWwkQ2VpeYKrewkbt4CWoqxI7ZDcqzVZQBoIm7sAgAAgO3QiQVsMkYwP7BA/UfMltMVovSwLKtLAgDAUoRYwMuxhQAAgMMRYgEvxhYCAADqR4gFvLgDG9V/i4ak57OFAACAn+HGLsCLhTlD1Csh1eoyAADwOoRYwIvnYE+IXGN1KQAAeCXGCQCvnoPNYg7Wx9TUuLV/f5XCwkLN/aQAAB8Psffee6/+/e9/a8WKFQoJCVFhYaHVJQGts4Ugd9F/DzNgDtaXrP9+ux5/+H29+/Yi7d9/QHFxYbr4ssG6acIoxcVHWF0eANiObUJsVVWVxo4dq9NPP13PPfec1eUALYotBL7t62UbdOG5f1d0hEeTbopU5nHBWrqiQi+88LE++mCJ3p99pxITo6wuEwBsxTYh9q9//av5cvr06VaXArToIQazEvPM7mtMhJMtBD7I7Xbr+qufVLfsAM2ekayI8J9uRbj4/AjdcGW0Bpy7TX/+40ua9sKNVpcKALZimxDbHJWVlebjoOLiYkvrAX4+PlCYs8bsvqbFGuMDHGLgixZ8vlobftil599NqQ2wB2WmB+sPv4nSbfd+pYLdxYpPiLSsTgCwG5/eTjB58mRFRUXVPlJTWVUE77HdVaHMtHjz5i0CrO9atXKzXK4A9evTrt7rw4eE6cABtzkzCwCwSYi97bbbzLtzj/RYu3Zts9//pEmTVFRUVPvYsmVLi9YPAEfTzhmiqiqPyso99V7fW1hjvgxtF9zGlQGAvVk6TnDLLbfoiiuuOOLbZGZmNvv9h4aGmg/AW7cQxDidygwfZnVJaEXDhp+s2255Uf96vVi/GR992PXnXi1ScsdondTjOEvqAwC7sjTEJiQkmA/AX7CFwP+kpMZr7C/66o/3LFJCfKAuPDtcAQEO7d/v1uPPFurFGSWa/MDlCgoKtLpUALAV29zYtXnzZu3du9d8WVNTY+6LNWRnZys8PNzq8oAjYguBf3vwsatUXFyuX167XGkdQ3RcWpBWra3S3n3VunniObr6ujOtLhEAbMfh8XjqH9TyMsbYwYsvvnjY859++qkGDx7cqPdhbCcwbvB6Y9rHcjnDWqFKoOEtBP3MLQTR3MTlp4xPtV8v3aA3ZizU3r2lSkuL1yW/GqTMrPZWlwYAXqWkuFwZHa8172eKjIy0f4htCYRYtHX3dX5ggfoPmq342HD1SkhlfAAAgBYKsbYZJwDsxgiwUb22qEtKIt1XAABaGCEWaEWpidFKp+kPAECL8+nDDgBLtxAcv0nZzpVWlwIAgE+iEwu06hYCdsACANAaCLFAC28hMHbApsXGMQcLAEArIsQCLbWFYMTBLQRZbCEAAKCVEWKBljhCtuciZSexhQAAgLZCiAWaiSNkAQCwDiEWOAaZafEakBpEgIUlyssr9f67X2nd2m1yhYXq7HNy1O34VKvLAoA2QYgFmjkHW5ZY9t8VWoQGtL2P/r1Mv71+qgoLy5WRFqq9hTW67563dPY5vfXktOsVHt7O6hIBoFWxJxZoxhzssu5LNLrPEsW5XHRh0eYWf/m9xl/2mAaeGqB1X6Rr/aI05X97nF6ckqR5c1fouiufsLpEAGh1dGKBRmILAbzFIw+8q+O7hGjGtPYKDnaYz4WEOHTZmEjz5cXXrdDyrzeqZ69Mq0sFgFZDiAUagS0E8BYlxeWa859v9eR9ibUB9lAXnh2u9onBeu+drwixAHwaIRY4CrYQwJuUllXK45E6dgis93pgoEPtk4JUUrK/zWsDgLZEiAWO1n2tPUJ2lNUlAYqPj1BUlFMLFldo1Jnhh10v2FOj1esqdeEl7S2pDwDaCjd2AQ3Y7qpQp66lGprZWWMyhlldDmAKDg7SxZcN0rSXSrTuh6o61zwej/5yf4EcjgBd9EtGXgD4NjqxwBHERrgk1Q0KgNVuuXW05v5nhfqfu003jI/UGf1d2llQrWkvFWvu/HI99NiViouPsLpMAGhVhFigoS0Eg2Yr3ml8u5Y9sPAuMbHh+uDjOzX5njf1yLT5uvvhvebzPXsdp5deO18jzu5tdYkA0OocHuP7T36iuLhYUVFRemPax3I5w6wuB16/hSCOLQTwemVlFdq+da95YlfHlDirywGAFtnCktHxWhUVFSkyMrLBt6MTC/wXWwhgR2Fh7dSpS7LVZQBAmyPEwu8Z4wOzEvPYQgAAgI0QYuHXzO5rd7qvAADYDSEWfh1go3pt0ZD0fLqvAADYDCEWfi01MVq9EqyuAgAANBWHHcBvtxCUxZUp27nS6lIAAEAz0ImFX28h6JWQxRwsAAA2RIiF32ALAQAAvoMQC7/bQpAWyyEGsI5xvsyBAzUKCeHTLwAcCz6LwqfRfYW32LRxp6Y8+oHeefMLlZRUqn37SF3yqyH6zU0jFR3DCYIA0FSEWPj0zVuFOWvovsJy337zoy4Yda+codX67VURysqI1pLlFZr2zw/0/szFen/WHYpPaPhoRQDA4Qix8FnbXRU6MS1eA1KDuHkLlo4P/PqqJ5SZJn3yeqqiowLN58ddFKmbr4nWwPO26y+TXtFTz15vdakAYCus2AKAVrRg/hp9v26HHrwztjbAHtQpM0R/uCFKM99ZpD0FJZbVCAB2RIiFT87B3le0WINy55h7YOnCwkqrvs2TyxWggac7670+4owwVVXV6Pt129q8NgCwM8YJ4FOYg4W3CQ0NUlWVR+X7PQpzOQ67Xlhc89PbtQu2oDoAsC86sfCp7mvmiPc0rEu+BqRmEWDhFXKHnayaGo9eeqO43usv/F+xOnSI0kk9jmvz2gDAzujEwse6r4mEV3iVtPQEXTDmdN36t6/UPjFI550VJofDocpKt/75QpGee7VYd0++VEFBdedlAQBHRoiFrbuv8wML1H/EbMXHhnOELLzWw1OuUlFhmS688ltlHReqzPQgffNdlXYVHND1N47Qr284y+oSAcB2CLGwLSPARvXaoi4pdF/h3cLC2un/3vqDvly4Tm++vlD79pZqzMUJuvTyQerStaPV5QGALRFiYWupidFK57Aj2IAxQtC3f1fzAQA4dtzYBdvOwbqO32Su0AIAAP6HTixsPgebyhwsAAB+iBAL220hGJS8UdlJzMECAODPCLGwYfe1M91XAAD8HCEWXt99XZ+6RyN7LqL7CgAAahFi4bWmbFwlV84m8xCDAal0XwEAwP8QYuHVMtPiNSA1iAALAADqYMUWvHYO1sAKLQAAUB86sfDqOdg4VxxdWPiM/Px9mvn2Yu3bW6LUtASdd34fRUS6rC4LAGyJEAuv3ELQhy0E8CE1NW7d+adX9czTHys42KGEuGBty6/Un2/7l/7291/p8vFDrC4RAGyHEAuvwBYC+LJ77npd056apXtui9Ovx0UpKjJQW7cf0N8e2quJv31OkVFOjb7gNKvLBABbYSYWXtGBNQLs6D5LNDSzMwEWPmVPQYmm/vMj3XFLrP54U6wZYA0pycGa+mCiRg0L0/33viGPx2N1qQBgK4RYeAVneKjSYqMZH4DP+ejfy1RdXaPrx0Ufds3hcOjGK6O0/vud+m7VZkvqAwC7IsTCK+Zgh3SaZ3UpQKsoKiqXyxmo+LifOrA/l9Yx2HxZXLS/jSsDAHsjxMLSOdhZiXkalDtH2UlxjBHAJ2VkJqm0rEbffFdZ7/WFS/bL4ZDSj0to89oAwM64sQuWdV85jQv+4MzhPdShQ5T+fN8evfNCBwUFOWqv7d1Xo388UaShuSepY0qcpXUCgN0QYmHhFgKj+zrK6pKAVhUcHKQHH7ta4y55RP3P3abfXh2p7IwQLVleoUemFqmoNEj/uu8yq8sEANshxKLNTNm4iu4r/NLwET311vuTdN89b+hXN3xvPhcY6NDIUTn6812/UFZ2e6tLBADbIcSiTcYHjNnXkbmLFBPh1JgMuq/wP/36d9P7s+7Q9m17tHdvqTp0iFVcfITVZQGAbRFi0WYrtLh5C5CSO8aZDwDAsWE7AVrdhspSq0sAAAA+hk4s2mQOdnh6ntLDsqwuBwAA+AhCLFp3CwFzsAAAoBUQYtHi2EIAAABaGyEWLYYtBAAAoK0QYtFi4wOFOWvM7mtaLFsIAABA6yLEokU6sMb866Vd8tUrIYvxAQAA0OpYsYUW2wMb53IRYAEAQJsgxKJF5mCHdJpndSkAAMCPME6AY9tC0P2nLQTMwQIAgLZEiEWTsYUAAABYjRCLJmELAQAA8AaEWDS6+zo/sED9R8xWfGw4WwgAAIClCLFo/PhAz0XKTkqk+woAACzHdgI0eoVWdhLjAwAAwDsQYnFUGypLFROXb3UZAAAAtRgnwNHHCM79aQtBeliq1SUBAACYCLGoF1sIAACANyPEog62EAAAADsgxKJO93V96h62EAAAAK9HiMX/jpDN+ekI2QGpnem+AgAAr0aIhTlC4M6s1sU9izQkiSNkAQCA92PFFgAAAGzHFiH2xx9/1FVXXaWMjAw5nU5lZWXpzjvvVFVVldWl+cQc7LLuSzT8+AVWlwIAAOBb4wRr166V2+3W1KlTlZ2drVWrVumaa65RWVmZHnzwQavLsyW2EAAAADuzRYg966yzzMdBmZmZWrdunZ566qkjhtjKykrzcVBxcXGr12oHbCEAAAB2Z4txgvoUFRUpNjb2iG8zefJkRUVF1T5SUzlxythCYB5i0GeJhmZ2JsACAABbskUn9ud++OEHTZky5aijBJMmTdLEiRPrdGL9NcjWdl9zfzpCdkwGWwgAAIB9WdqJve222+RwOI74MOZhD7Vt2zZztGDs2LHmXOyRhIaGKjIyss7DX+dfjQB7sPs6JmOY1SUBAADYtxN7yy236Iorrjji2xjzrwdt375dQ4YMUd++fTVt2rQ2qNB3OMNDlRYbzc1bAADAJ1gaYhMSEsxHYxgdWCPA9u7dWy+88IICAmw7zmvJFoIhneZJirO6HAAAAP+ZiTUC7ODBg5Wenm7Owe7evbv2Wvv27S2tzT5bCOK4iQsAAPgMW4TYTz75xLyZy3ikpKTUuebxeCyry9u3ELhyNml0ep4GpHZmjAAAAPgUW3xP3pibNcJqfQ8cPj5wX9FiDcqdo2Fd8jWu2ygCLAAA8Dm26MSiCd3X7j91X9NiGR8AAAC+ixDrI93XWYl57IAFAAB+gxDrQyu0uHkLAAD4C1vMxOLINlSWWl0CAABAm6IT6yNbCIan5yk9LMvqcgAAANqEX4XYg9sMyveXyRdM3bRakSdv04COWzW64xmSWyopLre6LAAAgGYrKdlvvjzaFiqHx4/2VG3dulWpqalWlwEAAICj2LJly2HnA/htiHW73dq+fbsiIiLkcDiO+f0VFxebodj4Q46MjGyRGmF/fFygPnxc4Of4mEB9+LiQ2YEtKSlRcnKyAgIavn3Lr8YJjD+IIyX65jI+yPz1Aw0N4+MC9eHjAj/HxwTq4+8fF1FRUUd9G7YTAAAAwHYIsQAAALAdQuwxCA0N1Z133mm+BA7i4wL14eMCP8fHBOrDx0Xj+dWNXQAAAPANdGIBAABgO4RYAAAA2A4hFgAAALZDiAUAAIDtEGJbyI8//qirrrpKGRkZcjqdysrKMu8urKqqsro0WOjee+9V37595XK5FB0dbXU5sMiTTz6p4447Tu3atdOpp56qr776yuqSYKHPP/9c55xzjnkakXF65Lvvvmt1SfACkydP1imnnGKeKpqYmKjRo0dr3bp1Vpfl1QixLWTt2rXmsbZTp07Vd999p0ceeURPP/20br/9dqtLg4WML2LGjh2r66+/3upSYJEZM2Zo4sSJ5he1X3/9tXr06KHhw4dr165dVpcGi5SVlZkfB8YXN8BB8+bN0w033KBFixbpk08+0YEDBzRs2DDz4wX1Y8VWK3rggQf01FNPaePGjVaXAotNnz5dEyZMUGFhodWloI0ZnVeju/LEE0+Yrxtf7Brnot9000267bbbrC4PFjM6se+8847ZdQMOtXv3brMja4TbgQMHWl2OV6IT24qKiooUGxtrdRkALOzEL1u2TLm5ubXPBQQEmK9/+eWXltYGwPszhIEc0TBCbCv54YcfNGXKFF133XVWlwLAIgUFBaqpqVFSUlKd543Xd+zYYVldALyb8R0b47t3/fr104knnmh1OV6LEHsUxrf7jG/3HOlhzMMeatu2bTrrrLPMWchrrrnGstrhPR8TAAA0ljEbu2rVKr322mtWl+LVgqwuwNvdcsstuuKKK474NpmZmbU/3r59u4YMGWLekT5t2rQ2qBDe/jEB/xUfH6/AwEDt3LmzzvPG6+3bt7esLgDe68Ybb9QHH3xgbrFISUmxuhyvRog9ioSEBPPRGEYH1giwvXv31gsvvGDOvsG/Pybg30JCQszPB3PmzKm9ccf4NqHxuvEfFQAcZNxnb9zwadzo99lnn5krO3FkhNgWYgTYwYMHKz09XQ8++KB5V+FBdFz81+bNm7V3717zpTEbuWLFCvP57OxshYeHW10e2oCxXmvcuHHKyclRnz599Oijj5orc8aPH291abBIaWmped/EQZs2bTI/Nxg38KSlpVlaG6wdIXj11Vc1c+ZMc1fswbn5qKgoc/88DseKrRZcodTQf0r8EfsvY+zgxRdfPOz5Tz/91PyiB/7BWK9lrNwz/lM6+eST9fjjj5urt+CfjC6b8V27nzO+2DH+L4F/Mu6nqI/xnd2jjbD5K0IsAAAAbIehTQAAANgOIRYAAAC2Q4gFAACA7RBiAQAAYDuEWAAAANgOIRYAAAC2Q4gFAACA7RBiAQAAYDuEWAAAANgOIRYAmsA4/tE4HrKhR2FhofxBRUWF+WfRvXt3BQUFafTo0VaXBMDPEGIBoInOOuss5efn13m89dZb8ic1NTVyOp367W9/q9zcXKvLAeCHCLEA0EShoaFq3759nUdsbGydt5k+fbqio6P17rvvqlOnTmrXrp2GDx+uLVu21L7NXXfdpZNPPrn29aqqKmVnZ9fp6L7++uvKysoyf35cXJzGjBmj3bt31/4c422NX+NQgwcP1oQJE2pff+mll5STk6OIiAiz1ksuuUS7du2qvf7ZZ5/V+TX37dunk046SZdffrk8Hk+9fwZhYWF66qmndM0115jvEwDaGiEWAFpJeXm57r33Xv3rX//SwoULzZD4y1/+ssG3f+KJJ7Rz5846z3Xt2tUMxOvWrdPHH3+sH3/8UX/84x+bVMeBAwd0991365tvvjEDr/E+jFGA+pSWlmrkyJHKzMzU888/b4ZbAPBGQVYXAAC+ygiPRjA99dRTzddffPFFdevWTV999ZX69OlT52337t2re+65xwyof/nLX2qfNzqiB8XExJjdWONb+U1x5ZVX1v7YCKePP/64TjnlFDOwhoeH116rrKw0O70ul0szZswwZ10BwFvRiQWAVmKEQCMsHtpVNUYM1qxZc9jb/u1vf9OQIUPUv3//w67Nnz/fDJvGz92/f78eeuihOtcvvvhi8/rBh/H2h1q2bJnOOeccpaWlmSMFgwYNMp/fvHlznbe79NJLNWfOHPO6MTIBAN6MEAsAFlu/fr2effZZ3X///fVeN+ZZly9frtmzZ2vPnj165pln6lx/5JFHtGLFitqH8fYHlZWVmbO4kZGReuWVV7RkyRK98847tTO4h9qxY4d5g9rf//53rVy5slV+rwDQUgixANBKqqurtXTp0trXjblWYy7WGCk4lDFCcPXVV5s3ddXH2AJg3BxmbAG49tprzTB6KOPGKuPnHnwYb3/Q2rVrzeB73333acCAAWY3+NCbug713nvv6YILLjBv1ho/frxZPwB4KwaeAKCVBAcH66abbjJnUI3RghtvvFGnnXZanXnYH374wfy2vvGyPq+99pq5nSApKcns2D799NN1Oq1HY4wQhISEaMqUKfr1r3+tVatWmTd51efghgUj8BqzuMbLP//5zw2+79WrV5vdXGOet6SkxOwCGw7duAAArYUQCwCtxLhByuiyGiuttm3bZnZCn3vuuTpvY3y7/69//ethK7oOMuZnb731VnNrQXx8vEaMGKEHH3yw0TUkJCSY2w1uv/12M0z36tXL/Pnnnntugz/HWJ9lbCYw9uEahxiceOKJ9b6dscUgLy+v9vWePXuaLxtaywUALcnh4bMNALQ4Izgau1r95QQvAGhrzMQCAADAdgixAAAAsB3GCQAAAGA7dGIBAABgO4RYAAAA2A4hFgAAALZDiAUAAIDtEGIBAABgO4RYAAAA2A4hFgAAALZDiAUAAIDs5v8BMDYP4gQEouQAAAAASUVORK5CYII=",
|
||
"text/plain": [
|
||
"<Figure size 800x600 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"def plot_decision_boundary(clf, X, y, title):\n",
|
||
" h = .02 # step size in the mesh\n",
|
||
" x_min, x_max = X[:, 0].min() - .5, X[:, 0].max() + .5\n",
|
||
" y_min, y_max = X[:, 1].min() - .5, X[:, 1].max() + .5\n",
|
||
" xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
|
||
" np.arange(y_min, y_max, h))\n",
|
||
" Z = clf.predict(np.c_[xx.ravel(), yy.ravel()]) #clf.predict(...) — модель предсказывает метку класса для каждой точки на плоскости.\n",
|
||
" #Мы берём 2 признака (X[:, 0] и X[:, 1]), чтобы построить 2D график.\n",
|
||
" #Эта сетка из точек xx и yy покрывает всё пространство данных.\n",
|
||
" Z = Z.reshape(xx.shape)\n",
|
||
"\n",
|
||
" plt.figure(figsize=(8, 6))\n",
|
||
" plt.contourf(xx, yy, Z, alpha=0.4) #Рисуется граница решения: цветная область, показывающая, где модель думает, что данные принадлежат к одному или другому классу.\n",
|
||
" plt.scatter(X[:, 0], X[:, 1], c=y, edgecolors='k')\n",
|
||
" plt.title(title)\n",
|
||
" plt.xlabel(\"Признак 1\")\n",
|
||
" plt.ylabel(\"Признак 2\")\n",
|
||
" plt.show()\n",
|
||
"\n",
|
||
"plot_decision_boundary(clf, X_test, y_test, \"Граница по Iris\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "ee0b9f89-fccb-4074-8abc-e8ceef6964cb",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Интерпретация результатов (датасет Iris)**\n",
|
||
"\n",
|
||
"Наблюдаемые эффекты:\n",
|
||
"Модель успешно обучается различать классы цветов ириса на основе выбранных признаков (например, длина и ширина лепестков).\n",
|
||
"Метрики качества (точность, полнота, F1-оценка) показывают высокую эффективность модели, особенно для легко различимых классов.\n",
|
||
"График границы принятия решений позволяет визуально оценить, как модель разделяет пространство признаков между классами.\n",
|
||
"\n",
|
||
"Практическая значимость:\n",
|
||
"Датасет Iris часто используется как эталонный пример для изучения базовых методов классификации, так как он хорошо сбалансирован и легко интерпретируется.\n",
|
||
"Анализ на этом датасете помогает понять, как различные алгоритмы (например, стохастический градиентный спуск) работают на многоклассовых задачах.\n",
|
||
"Визуализация результатов способствует лучшему пониманию внутренней логики модели и помогает выявить возможные проблемы, такие как перекрытие классов."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "4fb1a4e2-e7f3-4d11-b11f-1cbd361466bf",
|
||
"metadata": {},
|
||
"source": [
|
||
"\n",
|
||
"Зачем нужна plot_decision_boundary(...)?\n",
|
||
"Эта функция рисует, как модель разделяет классы. То есть она показывает, где проходит граница между разными классами (например, между видами цветов в Iris), которую нашла моя модель. Некоторые комментарии оставлены в коде, для его лучшего понимания.\n",
|
||
"График:\n",
|
||
"Я получил визуальное подтверждение, как хорошо (или плохо) модель разделила данные:\n",
|
||
"1)Если граница проходит между группами точек разного цвета — модель сработала хорошо.\n",
|
||
"2)Если граница граница проходит между группами точек и точки переходят через границу, другими словами точки нечетка разделены по цветам, а есть .\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "8dacd035-d50b-40ca-8934-ec90af0876cb",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Использование датасета make_blobs** "
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "46193c3b-7193-469a-827e-fd89df74e60f",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Импорт библиотек**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 22,
|
||
"id": "1832eaa6-4518-43a7-aae9-e51b543041e3",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"import numpy as np\n",
|
||
"import matplotlib.pyplot as plt\n",
|
||
"from sklearn.datasets import load_iris, make_blobs\n",
|
||
"from sklearn.linear_model import SGDClassifier\n",
|
||
"from sklearn.model_selection import train_test_split\n",
|
||
"from sklearn.preprocessing import StandardScaler\n",
|
||
"from sklearn.metrics import classification_report, confusion_matrix, ConfusionMatrixDisplay"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "fb92fcd5-f57e-4d85-986a-01ffa346d183",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Генерация данных**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 23,
|
||
"id": "b2dc3b2e-33d0-42ae-8952-15748ee1d158",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"X, y = make_blobs(n_samples=300, centers=3, cluster_std=0.5, random_state=0) #Генерируется 300 точек (n_samples=300).\n",
|
||
" #Они распределены по 3 кластерам (centers=3).\n",
|
||
" #Размер (разброс) каждого кластера — cluster_std=0.5.\n",
|
||
" #random_state=0 нужен для повторяемости результата."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "b2c36822-e5e6-42ea-9684-f71d38e39a39",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Визуализация данных**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 29,
|
||
"id": "dccd10a6-868b-41bf-9239-6a1a5dbeffa2",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 800x600 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"plt.figure(figsize=(8, 6))\n",
|
||
"plt.scatter(X[:, 0], X[:, 1], c=y, cmap='viridis')\n",
|
||
"plt.title(\"три рапределенных кластера\")\n",
|
||
"plt.xlabel(\"Признак 1\")\n",
|
||
"plt.ylabel(\"Признак 2\")\n",
|
||
"plt.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "011b5d94-5bb9-4d09-91b9-558f6ae88343",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Масштабирование и обучение модели**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 25,
|
||
"id": "024cbf99-b553-4da6-9117-67265fe31266",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"scaler = StandardScaler()\n",
|
||
"X_scaled = scaler.fit_transform(X)\n",
|
||
"X_train, X_test, y_train, y_test = train_test_split(X_scaled, y, test_size=0.3, random_state=42, stratify=y) #Масштабируем данные, чтобы признаки имели единичное среднее\n",
|
||
" #и стандартное отклонение (улучшает работу моделей).\n",
|
||
" #Делим данные на обучение и тест (70% на обучение, 30% на тест).\n",
|
||
"clf = SGDClassifier(loss=\"log_loss\", max_iter=1000, tol=1e-3, random_state=42) \n",
|
||
"clf.fit(X_train, y_train)\n",
|
||
"y_pred = clf.predict(X_test)\n",
|
||
"#Создаётся классификатор на стохастическом градиентном спуске (SGD).\n",
|
||
"#Используется логистическая регрессия (loss=\"log_loss\").\n",
|
||
"#Модель обучается на X_train, y_train."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "c5798cf4-de72-4743-bde2-efc9ad8647b1",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Оценка качества модели**"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 26,
|
||
"id": "d7c23ba8-091e-4845-a099-0e57a82f06d1",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" 0 1.00 1.00 1.00 30\n",
|
||
" 1 1.00 1.00 1.00 30\n",
|
||
" 2 1.00 1.00 1.00 30\n",
|
||
"\n",
|
||
" accuracy 1.00 90\n",
|
||
" macro avg 1.00 1.00 1.00 90\n",
|
||
"weighted avg 1.00 1.00 1.00 90\n",
|
||
"\n"
|
||
]
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 640x480 with 2 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"print(classification_report(y_test, y_pred))\n",
|
||
"ConfusionMatrixDisplay.from_predictions(y_test, y_pred)\n",
|
||
"plt.title(\"Матрица на make_blobs\")\n",
|
||
"plt.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "b93253d6-29d0-41eb-a594-38f1c40fa88e",
|
||
"metadata": {},
|
||
"source": [
|
||
"Визуализация границы решений"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 28,
|
||
"id": "6d80a9a5-372e-48c9-ba0c-a34042f17e3c",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 800x600 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"plot_decision_boundary(clf, X_test, y_test, \"граница на make_blobs\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "03ea6cba-5f55-43e6-9a47-872ea09ad122",
|
||
"metadata": {},
|
||
"source": [
|
||
"#Оставил комментарии в коде, для его лучшего понимания"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "9bc23ceb-44bb-4a7c-bd0d-fb9ac47fb911",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Интерпретация результатов (синтетический датасет make_blobs)**\n",
|
||
"\n",
|
||
"Наблюдаемые эффекты:\n",
|
||
"Алгоритм эффективно обучается на синтетически сгенерированных кластерах, каждый из которых представляет отдельный класс.\n",
|
||
"Визуализация границ решений показывает, как модель проводит разделение между кластерами на основе обучающих признаков.\n",
|
||
"Границы между кластерами могут быть четкими или размытыми в зависимости от плотности, расположения и стандартного отклонения точек (параметр cluster_std).\n",
|
||
"\n",
|
||
"Практическая значимость:\n",
|
||
"Использование make_blobs позволяет наглядно продемонстрировать, как классификатор (например, SGD-классификатор или логистическая регрессия) обучается различать классы."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "6c8f9601-39fc-4902-9c21-6fec103a5903",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Использвание внешнего датасета \"Dataset_Malavi_National_Footbal_Team\"**\n",
|
||
"\n",
|
||
"\n",
|
||
"\n",
|
||
"Импорт библиотек"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 55,
|
||
"id": "93fdb140-5819-45e1-853d-534790c2777d",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"import pandas as pd\n",
|
||
"import numpy as np\n",
|
||
"import matplotlib.pyplot as plt\n",
|
||
"from sklearn.model_selection import train_test_split\n",
|
||
"from sklearn.preprocessing import StandardScaler\n",
|
||
"from sklearn.linear_model import SGDClassifier\n",
|
||
"from sklearn.metrics import classification_report, ConfusionMatrixDisplay"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "740341e9-fcff-4587-b441-c66f83be749e",
|
||
"metadata": {},
|
||
"source": [
|
||
"Загрузка и предварительный анализ данных"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 62,
|
||
"id": "6b8e01f3-4600-4232-a4ea-6389679d7c31",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>Date</th>\n",
|
||
" <th>Opponent</th>\n",
|
||
" <th>Team Score</th>\n",
|
||
" <th>Opponent Score</th>\n",
|
||
" <th>Result</th>\n",
|
||
" <th>Venue</th>\n",
|
||
" <th>Competition</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>~1957</td>\n",
|
||
" <td>Northern Rhodesia</td>\n",
|
||
" <td>0.0</td>\n",
|
||
" <td>5.0</td>\n",
|
||
" <td>Loss</td>\n",
|
||
" <td>Unknown</td>\n",
|
||
" <td>Friendly (First International)</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>1</th>\n",
|
||
" <td>~1962</td>\n",
|
||
" <td>Ghana</td>\n",
|
||
" <td>0.0</td>\n",
|
||
" <td>12.0</td>\n",
|
||
" <td>Loss</td>\n",
|
||
" <td>Unknown</td>\n",
|
||
" <td>Friendly</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>2</th>\n",
|
||
" <td>~1968</td>\n",
|
||
" <td>Botswana</td>\n",
|
||
" <td>8.0</td>\n",
|
||
" <td>1.0</td>\n",
|
||
" <td>Win</td>\n",
|
||
" <td>Unknown</td>\n",
|
||
" <td>Friendly</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>3</th>\n",
|
||
" <td>02/06/1996</td>\n",
|
||
" <td>South Africa</td>\n",
|
||
" <td>0.0</td>\n",
|
||
" <td>3.0</td>\n",
|
||
" <td>Loss</td>\n",
|
||
" <td>Away</td>\n",
|
||
" <td>World Cup Qualifier</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>4</th>\n",
|
||
" <td>07/07/1996</td>\n",
|
||
" <td>Zambia</td>\n",
|
||
" <td>1.0</td>\n",
|
||
" <td>1.0</td>\n",
|
||
" <td>Draw</td>\n",
|
||
" <td>Home</td>\n",
|
||
" <td>Friendly</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" Date Opponent Team Score Opponent Score Result Venue \\\n",
|
||
"0 ~1957 Northern Rhodesia 0.0 5.0 Loss Unknown \n",
|
||
"1 ~1962 Ghana 0.0 12.0 Loss Unknown \n",
|
||
"2 ~1968 Botswana 8.0 1.0 Win Unknown \n",
|
||
"3 02/06/1996 South Africa 0.0 3.0 Loss Away \n",
|
||
"4 07/07/1996 Zambia 1.0 1.0 Draw Home \n",
|
||
"\n",
|
||
" Competition \n",
|
||
"0 Friendly (First International) \n",
|
||
"1 Friendly \n",
|
||
"2 Friendly \n",
|
||
"3 World Cup Qualifier \n",
|
||
"4 Friendly "
|
||
]
|
||
},
|
||
"execution_count": 62,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Путь к файлу — если он в той же папке, где запущен notebook\n",
|
||
"df = pd.read_csv(\"Dataset_Malawi_National_Football_Team_Matches.csv\")\n",
|
||
"\n",
|
||
"# Посмотреть первые строки\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e156ee56-eb65-49a8-b1cd-e10e7e287967",
|
||
"metadata": {},
|
||
"source": [
|
||
"Подготовка признаков и целевой переменной"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 63,
|
||
"id": "099406fe-2b87-4bcc-9ea9-926e8772aa76",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stderr",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"C:\\Users\\futbo\\AppData\\Local\\Temp\\ipykernel_22688\\2875606193.py:1: UserWarning: Could not infer format, so each element will be parsed individually, falling back to `dateutil`. To ensure parsing is consistent and as-expected, please specify a format.\n",
|
||
" df[\"Date\"] = pd.to_datetime(df[\"Date\"], errors=\"coerce\")\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"df[\"Date\"] = pd.to_datetime(df[\"Date\"], errors=\"coerce\")\n",
|
||
"\n",
|
||
"# Удаляем строки с ошибочными датами\n",
|
||
"df = df.dropna(subset=[\"Date\"])\n",
|
||
"\n",
|
||
"# Дальше можно продолжать анализ как обычно\n",
|
||
"df[\"year\"] = df[\"Date\"].dt.year\n",
|
||
"# Удалим строки с пропущенными значениями в нужных колонках\n",
|
||
"df = df.dropna(subset=[\"Team Score\", \"Opponent Score\"])\n",
|
||
"\n",
|
||
"# Целевая переменная: победа Малави\n",
|
||
"df[\"Malawi_win\"] = (df[\"Team Score\"] > df[\"Opponent Score\"]).astype(int)\n",
|
||
"\n",
|
||
"# Признаки: домашняя игра, континент соперника, год, и т.п. — упростим\n",
|
||
"df[\"is_home\"] = df[\"Competition\"].str.lower().str.contains(\"home\").astype(int)\n",
|
||
"df[\"year\"] = pd.to_datetime(df[\"Date\"]).dt.year\n",
|
||
"\n",
|
||
"# Используем только числовые признаки\n",
|
||
"features = [\"Team Score\", \"Opponent Score\", \"is_home\", \"year\"]\n",
|
||
"X = df[features]\n",
|
||
"y = df[\"Malawi_win\"]"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "0bcde4d4-e4ef-426f-87d4-ebf4a1c0860f",
|
||
"metadata": {},
|
||
"source": [
|
||
"Разделение и масштабирование"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 58,
|
||
"id": "cf697d3a-6247-4a19-93cd-528ec3fd2c94",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"# Делим на тренировочную и тестовую выборки\n",
|
||
"X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)\n",
|
||
"\n",
|
||
"# Масштабирование признаков\n",
|
||
"scaler = StandardScaler()\n",
|
||
"X_train_scaled = scaler.fit_transform(X_train)\n",
|
||
"X_test_scaled = scaler.transform(X_test)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "101df8ae-bc8d-4322-a83a-8f7672b2dab3",
|
||
"metadata": {},
|
||
"source": [
|
||
"Обучение SGDClassifier"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 59,
|
||
"id": "cf493c09-b632-439c-8244-fffd56e4dcd3",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" 0 1.00 1.00 1.00 16\n",
|
||
" 1 1.00 1.00 1.00 2\n",
|
||
"\n",
|
||
" accuracy 1.00 18\n",
|
||
" macro avg 1.00 1.00 1.00 18\n",
|
||
"weighted avg 1.00 1.00 1.00 18\n",
|
||
"\n"
|
||
]
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 640x480 with 2 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"clf = SGDClassifier(loss=\"hinge\", alpha=0.01, max_iter=1000, tol=1e-3, random_state=42)\n",
|
||
"clf.fit(X_train_scaled, y_train)\n",
|
||
"\n",
|
||
"# Предсказания и отчёт\n",
|
||
"y_pred = clf.predict(X_test_scaled)\n",
|
||
"print(classification_report(y_test, y_pred))\n",
|
||
"\n",
|
||
"ConfusionMatrixDisplay.from_estimator(clf, X_test_scaled, y_test)\n",
|
||
"plt.title(\"Confusion Matrix - Победа Малави\")\n",
|
||
"plt.show()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "4fb5d352-b01e-4def-b2c4-62ca2116b041",
|
||
"metadata": {},
|
||
"source": [
|
||
"Визуализация границы решений (с 2 признаками)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 64,
|
||
"id": "29948e10-3aa9-49f5-8f94-4c07509b9b3d",
|
||
"metadata": {
|
||
"scrolled": true
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" 0 1.00 1.00 1.00 16\n",
|
||
" 1 1.00 1.00 1.00 2\n",
|
||
"\n",
|
||
" accuracy 1.00 18\n",
|
||
" macro avg 1.00 1.00 1.00 18\n",
|
||
"weighted avg 1.00 1.00 1.00 18\n",
|
||
"\n"
|
||
]
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 640x480 with 2 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 800x600 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"scaler = StandardScaler()\n",
|
||
"X_train_scaled = scaler.fit_transform(X_train)\n",
|
||
"X_test_scaled = scaler.transform(X_test)\n",
|
||
"clf = SGDClassifier(loss=\"hinge\", alpha=0.01, max_iter=1000, tol=1e-3, random_state=42)\n",
|
||
"clf.fit(X_train_scaled, y_train)\n",
|
||
"\n",
|
||
"# Предсказания и отчёт\n",
|
||
"y_pred = clf.predict(X_test_scaled)\n",
|
||
"print(classification_report(y_test, y_pred))\n",
|
||
"\n",
|
||
"ConfusionMatrixDisplay.from_estimator(clf, X_test_scaled, y_test)\n",
|
||
"plt.title(\"Confusion Matrix - Победа Малави\")\n",
|
||
"plt.show()\n",
|
||
"# Только два признака для визуализации\n",
|
||
"X_viz = df[[\"Team Score\", \"Opponent Score\"]].values\n",
|
||
"y_viz = df[\"Malawi_win\"].values\n",
|
||
"\n",
|
||
"# Масштабирование\n",
|
||
"scaler_viz = StandardScaler()\n",
|
||
"X_viz_scaled = scaler_viz.fit_transform(X_viz)\n",
|
||
"\n",
|
||
"# Обучим классификатор\n",
|
||
"clf_viz = SGDClassifier(loss=\"hinge\", alpha=0.01, max_iter=1000, tol=1e-3, random_state=42)\n",
|
||
"clf_viz.fit(X_viz_scaled, y_viz)\n",
|
||
"\n",
|
||
"# Построим границу решений\n",
|
||
"def plot_decision_boundary(clf, X, y, title):\n",
|
||
" h = .02\n",
|
||
" x_min, x_max = X[:, 0].min() - .5, X[:, 0].max() + .5\n",
|
||
" y_min, y_max = X[:, 1].min() - .5, X[:, 1].max() + .5\n",
|
||
" xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
|
||
" np.arange(y_min, y_max, h))\n",
|
||
" Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])\n",
|
||
" Z = Z.reshape(xx.shape)\n",
|
||
"\n",
|
||
" plt.figure(figsize=(8, 6))\n",
|
||
" plt.contourf(xx, yy, Z, alpha=0.4, cmap='coolwarm')\n",
|
||
" plt.scatter(X[:, 0], X[:, 1], c=y, edgecolors='k', cmap='coolwarm')\n",
|
||
" plt.title(title)\n",
|
||
" plt.xlabel(\"Team Score (scaled)\")\n",
|
||
" plt.ylabel(\"Opponent Score (scaled)\")\n",
|
||
" plt.show()\n",
|
||
"\n",
|
||
"plot_decision_boundary(clf_viz, X_viz_scaled, y_viz, \"Граница решений SGD (Malawi vs Opponent)\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "9a81f724-6b66-46b9-b4c1-abee793ed952",
|
||
"metadata": {},
|
||
"source": [
|
||
"**Интерпретация результатов**\n",
|
||
"Наблюдаемые эффекты:\n",
|
||
"Модель на основе SGDClassifier построила линейную границу, отделяющую победы Малави от остальных исходов.\n",
|
||
"Классификация возможна даже с небольшим числом признаков, таких как счёт и место проведения.\n",
|
||
"Погрешности возможны из-за ничей, нестандартных условий матчей и других скрытых факторов.\n",
|
||
"\n",
|
||
"Практическая значимость:\n",
|
||
"Такой подход может быть полезен для предварительного прогнозирования исходов матчей или анализа исторических результатов.\n",
|
||
"Метод подходит как часть большей модели предсказания, если использовать больше признаков (соперник, турнир, форма команды и т.п.)."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "b4b5ab86-54e2-4a95-ad29-0508750b6361",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": []
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "1bd31a25-3cbc-4dee-a9c1-2b132600ed40",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": []
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "Python 3 (ipykernel)",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"language_info": {
|
||
"codemirror_mode": {
|
||
"name": "ipython",
|
||
"version": 3
|
||
},
|
||
"file_extension": ".py",
|
||
"mimetype": "text/x-python",
|
||
"name": "python",
|
||
"nbconvert_exporter": "python",
|
||
"pygments_lexer": "ipython3",
|
||
"version": "3.13.3"
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 5
|
||
}
|