03_Classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyNjcXXiEzNHJAaUhVX+n32P",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/sim-so/HandsOnML/blob/main/03_Classification.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 분류 Classification\n",
        "1. scikit-learn 라이브러리 사용법 익히기\n",
        "2. 이진 분류 / 다중 분류 수행하기\n",
        "3. 모델의 성과 분석하기\n",
        "- 실습에 사용한 모델들이 어떤 원리로 분류를 수행하는지는 5~7장에 걸쳐 공부할 예정이므로,\\\n",
        "이번 장에서는 전반적인 머신러닝 수행과정에 초점을 두고 실습을 진행해봅시다."
      ],
      "metadata": {
        "id": "a9QIuKUJQN3t"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Load Library"
      ],
      "metadata": {
        "id": "CBEaOOurQ6nJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "\n",
        "import matplotlib as mpl\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns"
      ],
      "metadata": {
        "id": "58h2BAPbOB8j"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Functions\n",
        "데이터 시각화를 위한 함수를 사전 정의합니다."
      ],
      "metadata": {
        "id": "YevzEuYGGqhN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_digit(data):\n",
        "    image = data.reshape(28, 28)\n",
        "    plt.imshow(image, cmap=mpl.cm.binary, interpolation=\"nearest\")\n",
        "    plt.axis(\"off\")"
      ],
      "metadata": {
        "id": "91ZqPPoMp3Gh"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_digits(instances, images_per_row=10, **options):\n",
        "    size = 28\n",
        "    image_per_row = min(len(instances), images_per_row)\n",
        "    images = [instance.reshape(size, size) for instance in instances]\n",
        "    n_rows = (len(instances) - 1) // images_per_row + 1\n",
        "    row_images = []\n",
        "    n_empty = n_rows * images_per_row - len(instances)\n",
        "    images.append(np.zeros((size, size * n_empty)))\n",
        "    for row in range(n_rows):\n",
        "        rimages = images[row * images_per_row : (row + 1) * images_per_row]\n",
        "        row_images.append(np.concatenate(rimages, axis=1))\n",
        "    image = np.concatenate(row_images, axis=0)\n",
        "    plt.imshow(image, cmap = mpl.cm.binary, **options)\n",
        "    plt.axis(\"off\")"
      ],
      "metadata": {
        "id": "_gIo8iozVlNN"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1. MNIST\n",
        "MNIST 데이터셋은 가장 기본적인 이미지 데이터셋입니다.\\\n",
        "0에서 9까지 숫자들의 손글씨 이미지로 크기는 784 = (28, 28), 색상은 흑백입니다.\\\n",
        "7만개의 데이터 중 마지막 1만개는 테스트 데이터로서 사용합니다."
      ],
      "metadata": {
        "id": "YRDy2SAIr-qA"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 데이터셋 다운로드"
      ],
      "metadata": {
        "id": "Z0leLs3wHHTu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.datasets import fetch_openml\n",
        "\n",
        "mnist = fetch_openml('mnist_784', version=1, as_frame=False)        # default: as_frame=True\n",
        "mnist.keys()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DH94jXJ1LjhQ",
        "outputId": "3fc96104-1301-46bf-9ded-ea443a7d90db"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dict_keys(['data', 'target', 'frame', 'categories', 'feature_names', 'target_names', 'DESCR', 'details', 'url'])"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 이미지 데이터 확인"
      ],
      "metadata": {
        "id": "7G9CX8kpHKID"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X, y = mnist['data'], mnist['target']\n",
        "print(X.shape, y.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J29kXUENLubD",
        "outputId": "714348a0-ec18-477e-cdbe-cd8af12c75ae"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(70000, 784) (70000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "some_digit = X[0]\n",
        "some_digit_image = some_digit.reshape(28, 28)\n",
        "\n",
        "plt.imshow(some_digit_image, cmap='binary')\n",
        "plt.axis(\"off\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "id": "xDXTGbEsMtAC",
        "outputId": "21876ccc-042d-4b24-a529-e44e13a89d3e"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAADnCAYAAADl9EEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAGaElEQVR4nO3dPUiWfR/G8dveSyprs2gOXHqhcAh6hZqsNRqiJoPKRYnAoTGorWyLpqhFcmgpEmqIIByKXiAHIaKhFrGghiJ81ucBr991Z/Z4XPr5jB6cXSfVtxP6c2rb9PT0P0CeJfN9A8DMxAmhxAmhxAmhxAmhljXZ/Vcu/H1tM33RkxNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCiRNCLZvvG+B//fr1q9y/fPnyVz9/aGio4fb9+/fy2vHx8XK/ceNGuQ8MDDTc7t69W167atWqcr948WK5X7p0qdzngycnhBInhBInhBInhBInhBInhBInhHLOOYMPHz6U+48fP8r92bNn5f706dOG29TUVHnt8PBwuc+nLVu2lPv58+fLfWRkpOG2du3a8tpt27aV+759+8o9kScnhBInhBInhBInhBInhBInhGqbnp6u9nJsVS9evCj3gwcPlvvffm0r1dKlS8v91q1b5d7e3j7rz960aVO5b9iwody3bt0668/+P2ib6YuenBBKnBBKnBBKnBBKnBBKnBBKnBBqUZ5zTk5Olnt3d3e5T0xMzOXtzKlm997sPPDx48cNtxUrVpTXLtbz3zngnBNaiTghlDghlDghlDghlDghlDgh1KL81pgbN24s96tXr5b7/fv3y33Hjh3l3tfXV+6V7du3l/vo6Gi5N3un8s2bNw23a9euldcytzw5IZQ4IZQ4IZQ4IZQ4IZQ4IZQ4IdSifJ/zT339+rXcm/24ut7e3obbzZs3y2tv375d7idOnCh3InmfE1qJOCGUOCGUOCGUOCGUOCGUOCHUonyf80+tW7fuj65fv379rK9tdg56/Pjxcl+yxL/HrcKfFIQSJ4QSJ4QSJ4QSJ4QSJ4Tyytg8+PbtW8Otp6envPbJkyfl/uDBg3I/fPhwuTMvvDIGrUScEEqcEEqcEEqcEEqcEEqcEMo5Z5iJiYly37lzZ7l3dHSU+4EDB8p9165dDbezZ8+W17a1zXhcR3POOaGViBNCiRNCiRNCiRNCiRNCiRNCOedsMSMjI+V++vTpcm/24wsrly9fLveTJ0+We2dn56w/e4FzzgmtRJwQSpwQSpwQSpwQSpwQSpwQyjnnAvP69ety7+/vL/fR0dFZf/aZM2fKfXBwsNw3b948689ucc45oZWIE0KJE0KJE0KJE0KJE0KJE0I551xkpqamyv3+/fsNt1OnTpXXNvm79M+hQ4fK/dGjR+W+gDnnhFYiTgglTgglTgglTgglTgjlKIV/beXKleX+8+fPcl++fHm5P3z4sOG2f//+8toW5ygFWok4IZQ4IZQ4IZQ4IZQ4IZQ4IdSy+b4B5tarV6/KfXh4uNzHxsYabs3OMZvp6uoq97179/7Rr7/QeHJCKHFCKHFCKHFCKHFCKHFCKHFCKOecYcbHx8v9+vXr5X7v3r1y//Tp02/f07+1bFn916mzs7PclyzxrPhvfjcglDghlDghlDghlDghlDghlDghlHPOv6DZWeKdO3cabkNDQ+W179+/n80tzYndu3eX++DgYLkfPXp0Lm9nwfPkhFDihFDihFDihFDihFDihFCOUmbw+fPncn/79m25nzt3rtzfvXv32/c0V7q7u8v9woULDbdjx46V13rla2753YRQ4oRQ4oRQ4oRQ4oRQ4oRQ4oRQC/acc3JysuHW29tbXvvy5ctyn5iYmNU9zYU9e/aUe39/f7kfOXKk3FevXv3b98Tf4ckJocQJocQJocQJocQJocQJocQJoWLPOZ8/f17uV65cKfexsbGG28ePH2d1T3NlzZo1Dbe+vr7y2mbffrK9vX1W90QeT04IJU4IJU4IJU4IJU4IJU4IJU4IFXvOOTIy8kf7n+jq6ir3np6ecl+6dGm5DwwMNNw6OjrKa1k8PDkhlDghlDghlDghlDghlDghlDghVNv09HS1lyMwJ9pm+qInJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4QSJ4Rq9iMAZ/yWfcDf58kJocQJocQJocQJocQJocQJof4DO14Dhyk10VwAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "WWRekiVWNII8",
        "outputId": "fdcdb161-7236-478d-edb0-b236fbffd423"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'5'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_digits(X[:25], images_per_row=5)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "id": "hCi8R3v_G7b0",
        "outputId": "215e0add-3063-4fc4-8235-e22f531d3c09"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAADnCAYAAADl9EEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydeXCb95nfPzhJAiABAiAJniApnuJN6j5s2ZZV106c2E6yrmxvjvUkm+022Wy7md1ON5vsTDvtdCftbo+kbZJ67cTrKInj2o7kRHJ0WKREShQP8RIv8ARBkARI3Df6h+Z9V5Qp25IIgM7yM8N/CJD4vS/e53c8x/eRxONxttlmm62HNNUD2GabbTZm2zi32WaLsm2c22yzRdk2zm222aJsG+c222xR5B/y+rYrd5ttEo9ko19ur5zbbLNF2TbObbbZomwb5zbbbFG2jXObbbYo28a5zTZblA/z1m4DRCIRLBYLQ0ND9Pf3895772EwGGhsbOThhx+msLCQoqKiVA9zm98xkmqc0WiUWxPto9EoS0tLxGIxYrEYgUCArKwsjEYj8/PzeDwe1tbWkEqlpKWlkZubi8FgQKVSJW3MoVAIl8vFlStX6Ovro6+vj4sXL2I0GnG73ahUKurq6j5Wxunz+fB6vUxPT5OXl0dOTg7p6empHpZIOBxmcXGRlZUV3G43ACaTiby8PDQaDRLJhpGH3zkkH1KVsmlxzmg0itfrJRKJiL9zuVy8+uqr+P1+gsEgFouFgwcP8pnPfIbvfve7dHd3c+HCBZRKJQUFBXz+85/n6aefprGxcbOG9aHMzc0xMjLCF77wBRwOB8FgcN3rmZmZPPzww7z++utJG9P9MjAwQH9/P3/+53/Oiy++yPHjxykrK0Mmk6V6aADYbDZ+9KMf8ctf/pLu7m7i8Thf/OIX+dKXvsSePXtQKpWpHuJms+Fss6krp8/nw+fzEY1GcTqdLC4uYrFY8Pl8BAIBrl+/ztramvj+cDjMzMwM0WiUaDSKUqkkFovhdrs5e/YsNpuN9PR0TCYTlZWV7Ny5E61Wu5lDviOhUIjZ2VneeOMNLl68yOrqqjixpKWlIZfLkUqlBINBnE4nY2Nj5OXloVar7+shF1YMr9dLfn5+QlbkxcVFrFYrZrMZnU636f//fujv76e/v5833niD6elpJBIJ/1TLGjfNOEOhEAsLCywsLBAKhbDb7czMzDA0NITL5SIQCNDZ2YnT6dx4IHI51dXVuFwuxsfH8fl8KJVKzGYzpaWlVFRUUFJSglqt3qwhb0g8HiccDrO2tsbQ0BCdnZ10dHTg9/uJx+NIJBIyMjJQq9VoNBpmZmZYXV2lt7eX+vp68vLy0Ov19/z5TqeT6elpVlZWkMvlCTHO1dVVHA4HGo0GhUKx6f//fpifn2dsbIyhoSHC4TAAEokkZVvZUChEKBQSd3fRaFQcl1wuR6fTiZNxMBhEqVSSlpaGUqm87zFvinGGQiFGRkb4/ve/zy9+8QtcLhfxeHzdD9zc2m6E8MC/+OKLGAwG0tLSeOKJJ0hLS0Ov15OZmUlmZiZlZWVIpYlzMMdiMYLBIBMTEwwODvKtb30Lu90uXg+ATCajqKiIlpYWjh07xl/91V8xMDDACy+8wPHjxzl06BBf+MIX7nmc09PTdHV1MT09jVKppKWlZTMvkXg8jsPhwGq1srCwgNfr3VIr08rKCna7nVAolPJxxWIxRkZGGBkZoaOjg76+Pmw2G1NTUwDk5eXx9a9/Hb1ej0KhoLu7m+rqatra2qivryctLe2+Pn9TjFMqlaLValGr1UilUiKRyB0NMS0tDZPJJG5hZ2ZmkEgkpKen09bWhtFoFGciuVxOeno6SqUShUKRUMOEm2edM2fO0N3dzeTkJHa7nUAgsO4hicfj2Gw2nE4nsViMwsJCAoEA8/PzLC0tsbi4eF9jmJ6e5tq1a2RmZm76wxmNRvF4PAwODtLf309eXp7oEEr0vf0wfD4f8/PzdHV10dPTQzweR6PRkJOTw9NPP83evXspKSlJ2rnYbrczOzvLa6+9xtTUFOPj4ywtLRGNRikuLiYQCCCRSHjrrbdIS0tDJpOJO0ev18uOHTu2lnFmZ2djMBhwu93EYjFkMhk+n2/dqqNSqaiqqkKlUokPSzQaRaPRUF1dTU5OzmYM6a6JxWLYbDZOnjzJxYsXsVqtwM0JQqlUIpfLiUQihMNhHA6HuJoWFxfj9/uZm5vD5XLhdDrvy6hsNhujo6M0Nzdv1qWJCF7Q0dFRJiYmqK+vJycnJ+Xnzmg0yurqKkNDQ/T29jIyMgKATqdjx44d/MEf/AEGg+G+z/N3Mx6r1crVq1d58803sdvt+P1+lEoler2enTt34nK5WF1d5dKlS0SjUdEfEQwGSUtLIxQK3fc4NtU4H3vsMQoKCrhw4QJZWVnk5ubyt3/7tywtLQFQUlLCrl27+O53v0tGRgaxWIzR0VE6OzsZHR2975nmfhgfH6enp4dz586tc1qVlpZSXFzMoUOHuHjxIufOnUOv11NcXMzOnTvJzMykq6uLrq6udVv4eyUejxOLxe73cjZkenqav/zLv2RgYACZTEZJSUnSHGwfxNDQEJcuXeK//Jf/gtVqxefzAfDggw9y9OhRSkpKNuUM92HEYjFCoRDDw8O89tprvPzyyzidTrKzs2loaOD48eNUV1fT0NBAOBxmdnaWb33rWwwPDzM7OwuA0Wikrq5uUzzKm+YQkkgk5OfnI5FIUCqVqFQqdDodZ8+eZWxsjJmZGWpra2loaMBoNKJQKIjH4+zYsQMAs9mcEuMMBAK4XC5OnTrFlStXcLlcRCIRcft9+PBhGhsbyczMZHV1lXA4TEtLCy0tLeTn56NSqcTJZ3FxkenpaZaWltDpdHcVOxRWj7W1NdH5tNmEw2GWl5dFx0VpaSmZmZmb/jl3QzweZ3R0lJGREWw2G4FAALlcTm5urvi8KBSKpDiEbDYb8/Pz/OIXv6Czs5PV1VVyc3Opq6vj2LFjtLW1kZ+fj16vZ2ZmhuXlZRYWFvD5fEgkEtRqNUVFRdTX128t4wTIz88nPz+fyspK5HI5crmc/fv3I5FImJmZobm5mV27dq0zQuHckyo8Hg/T09O8/PLLjI6OinHM9PR0Ghoa+PSnP83Bgwe5evUqMpmM0tJSnnrqKYxGIxqNhoKCAubn58VrVCqVTE1NsWPHjrsyzkgkwtzcnBh4T4RxxmIx/H4/kUgEtVpNdXU1BoNh0z/nbsfU19fHwMCAuGPJysqiqqqK5uZmmpqakjKOaDTK5OQkXV1d/I//8T/ESaKyspIjR47w4osvij4VAIvFQl9fHyMjI0QiEWQyGXq9noqKCnbv3r0pSR0JyRBSqVRIJBJisRhqtZqMjAwALly4QDgc5tixYyl3QMTjcebn5+np6eHXv/418/PzRCIRdDodWq2WoqIinnzySSorK9HpdBw8eFB0dGVmZm549hFCSC+//DLHjx+/q/Ozz+ejvb2dmZmZ9yU6bAYul4ulpSWsVit+v5+srCxycnKSmm11O0L21enTpxkcHBRXn9LSUl588UWqq6uTMo6VlRVeffVVTp06RU9PD4FAgNraWnbt2sVXvvIViouL1xkmwOnTp3nnnXfEI4hareZP/uRP2LdvHzqdblOe74QYpzAwiURCRUUFi4uL5OTkYLfbsVgs2O32u972bSZCLLO/v5+enh76+/sJBoPodDpaWlowmUyYTCbKysrIyspCKpV+pPhqPB4nFAphtVrxer13NaZIJILNZhP/TqPRbOo2f2xsjOHhYVwuFwqFAo1GQ0ZGBnJ5atKr/X4/s7OzDAwMYLPZ8Hg8yGQyCgsLqaqqora2Ninn4bm5OSYmJrh48SJjY2M4HA5ycnLYuXMnBw8epLy8HK1Wi1QqFc+kTqeTubk5FhYWiMVi5ObmUlpaSnNzM0VFRZvmtEroNyOVSnn88cfRarX09vbS09PD9evX6evro7Gxkfz8/ER+/B0RtncnTpygr6+P/v5+1Go1VVVVfP3rX6eqqorMzEyCweBdPyCRSITl5WUCgcBd/V04HGZ+fh632y3GUjfTi3rq1Ckx0yk/P5/c3NyUGWc8Hmd5eZlz587xve99b134ae/evRw8eJCGhoaknDPfe+892tvbef3118V4e2trK4899hjPPvssaWlp4jgikQgOh4OrV68yPT3N6uoq8XiclpYWHnroIfbt27epC07Cv5m0tDTq6ur4N//m3/Dtb3+bmZkZ/uZv/oa6ujqKi4spKyujvr6eqqqqRA9FZGVlhampKTo7O7FarUilUv74j/+YXbt2sWfPHjIyMpDJZMTj8ZQ8vMJZx2g03tPfRyIRAoEAAwMDTE1N0d/fzzvvvMPc3Bxw0wv6wAMPoNVqU5KnGo1G6e3tpb+/H4vFQjgcRqvVUlVVxeOPP86uXbsSbpjRaBS/3097ezvvvvsusViMHTt2UFVVxbe//W2KiopED7Hf72dlZYUf/ehHDA8PMzQ0JPoX9u3bx5NPPsnRo0c3/V4m/MmTSqVkZ2fT1NREbW0tfr+fwcFB0REzPz+PVColMzMTvV6PXC5PeCzLbrczODjI8vIyoVCIrKwsWltbaWpqIjs7+57+p5DaBzdX5vt16GRkZGz4ZQcCAfGc4/V6CYfDhMNh4vG4mNMcCoXEXGYhcX9hYUF0uBQXF1NZWZk0L+it+P1+1tbW6OvrY3JyEq/XS0ZGBvn5+ezevZvKysqk7KiEo83i4iILCwvAzfvS2tpKdXU1sViMhYUFMZ65sLBAe3s7o6OjzMzMAKDX66mpqaGiooLi4uJN96MkZVlIT0+nsLCQF198kc7OTv79v//3XLt2TcxRHBkZYWpqik996lMYDAaysrISOp7Lly/zgx/8ALfbjV6vp7a2lubmZjGsc7cIhrgZcU7h/7jd7g0dQ3Nzc2LKXX9/P1arlcXFRSKRCD6fj7ffflvMakpLS8NoNFJRUYFcLkehUBCJRCgoKKC8vDwl+apTU1P09fXxgx/8gOXlZQAqKys5fPgwf/EXf0F2dnZSVnPBON1ut1iWdvToUZ577jn8fj/9/f1cuXKFs2fPYrVamZubw+fzEYlExPum0Wg4duzYpmQDbURS92w1NTWig+XChQuMjIwwNjZGf38/TqeTaDRKU1MTbW1t4iF8MwmHwwwPDzM8PMz09DSxWIycnBza2truy2spfFkSiQSFQoHZbL7r+KFEIhGrXQKBAK+88grnzp2joKBg3fuEooBYLIbL5SIYDBKJRNBqtahUKvbu3YtWq8VgMFBRUYHBYCA3N5e/+qu/wu12I5VK0ev15ObmJtU4Y7EYTqeTM2fOcOLECVZWVsQsGoPBgNFoRKvVJu0YIZFIkMvlqFQqNBoNHo+HEydOcOXKFWQyGaurqywvL7O4uEgsFiM9PR2ZTEYkEsHj8aBSqdDr9bS2tt7z8ePDSKpx5uTkkJmZSUZGBuFwGKlUytraGmtraywvL4sPYk5ODtXV1ZvuzY1Go8zPz2O321ldXUUikaDX6+/5s4QVzu12i8al1WrZsWPHXa/+crlczHO12WwMDQ0xNzf3vv8zPT0tpogJxpyVlYVMJiMrK4va2lry8vIoLCykoaEBlUqFXC4Xi5TlcrlYUZNMhHt//fp12tvbxd/LZDIx1p1M770wkZpMJgoKCpicnKS/v5++vj4xl1vwaqvVanQ6nRiD9ng8aDQacnNzKSwsTNiEknRvR3p6OpWVlXz1q1/l937v9zh37hw///nPOXXqFG+99RY9PT2cP3+e//7f/zslJSWb+tnxeJxgMEgoFCISiYiG+dhjj931SheLxQiHw/zkJz+ho6MDgOrqalpaWviTP/mTuzbOnJwc/u2//bfs2bOHy5cv09fXRzQafd/qlpeXh8FgYP/+/RQVFZGbm0tDQwNSqVQsrRJ+pFIpFouFs2fPYrFYCAQCmEwmMe6cTDweDy+99BJ9fX3rrkkmk/HMM8/Q2tqa1PHIZDIyMzP5oz/6Iw4cOMC//tf/GpfLRTQapaCggJqaGurq6njyySfJzs5Go9Hwne98h56eHux2O21tbRw4cCChu4+UBLkkEgkqlQqpVMrevXu5dOmSWM3icDi4ceMGExMTpKWlbXr20K3nQiER/9aavI+CsAr09/dz+vRpJiYm0Gq1tLa2smfPHjIzM+9pNlUoFOzcuVOMt94pxzYjI4OioiIyMzNRqVQfaGwej4cbN27g8/lQqVQ0NTUl/Ex/O6urq6JjamlpSTyuZGdni/pL91MDez8IK99f/uVfEgqFiMViaLVa9Ho9OTk5VFRUiI62mZkZFhYWiMfjZGVlodfrf/eMU9AMksvlYvK1cJFerxer1cr8/LyoG5Mo0tLSSE9P/8gF3EJSusfjEVek7u5u3G63WOPZ0tJyX0naZrMZs9l8T3+7EYJWUDAYRK1WU19fn/QqFLvdzsTEBBMTE2KxvUwmw2g00tzcjMlkSvo2WyA7O5vs7Gxqamru+B673c7KygpWqxW73Y5EIkGn0yV8Qkm6cfr9fqampuju7sZisTA8PExvb69YcpOVlUVJSQltbW1UVFQkbBxyuZxDhw6xc+fOj/w3y8vLzM/P85Of/IS+vj46Oztpbm5m3759fPOb3yQjIyMl4YmPik6n4+jRo0nPZf7BD37Az372MzGjRiqVUltbyyc/+Um++tWvpqxM8KMSjUZFz7nw3ba1tfHAAw8kNA01KcYZDofxer1MTEywsLBAb28vY2NjLCwsYLVaRZe64FTJysoiLS0toRIaQv3mnWRTbh//yMgIQ0NDDAwM0NnZidvtprCwkEcffZS2traUJ5B/FIT7m+y85rW1NbFQGW6umo2NjVRWVmIwGLaMsNidUKvVFBcXi+ESofIq0VVUCTfOeDyOz+fDarVy6tQpBgcHOX/+PKurq+tS3ATvmUajEZMREjWeeDxOJBJhcHCQ5uZmYrHYOp2a2yVWvF4vv/3tbzl79iznz5/H6/VSVlbGwYMHef755yktLU3IWBNBsqU/hEwcoawKbp6t9+/fT21t7cdCSS8rKysh0YMPI2HGGQgEcLvdXLhwgcHBQc6ePcvk5CQul0sM5grk5eVRVFTE7/3e77Fz504aGhowmUwJGdetRuh0Omlvb+c//af/xMMPP4zRaEStVmOxWJibm+P8+fNYrVZWVlaYmJgAQKvVsmvXLhobG3n00UdT5si4WwRP9czMDMXFxUlZ6efn5/ne975Hb2/vuklBIpF8rLR+l5aWGBkZwePxJPVzN9U4BWUyr9fL3NwcY2NjvPfee4yNjTEyMiKmlsHN7AqVSoXZbKasrIzy8nJRJ6a4uHgzh3VHIpEICwsLXLp0CZlMhsFgIDMzk/HxcVGmwm63i1IqOTk55Ofn88ADD1BdXU15eXlK1RvuFkHA7E76TpuJ1+vFbrfT19eHw+EAbhqlwWCgoKCA7OzslJar3Q2BQAC73S5mtCWLTTVOv98vSlueOXOGn/3sZ0xNTW2YhmYymaioqOCLX/wijY2NVFVVJd2RIgiMzczMcPnyZTIyMlCpVMzPz68r+VIqlWJ6X01NDf/qX/2rLSHvcTfcmvebDGw2G+Pj4/T29ooJH0Ko6NChQ+Tl5X1sjNPv94uSr8nkvo0zEAiwtLTE3//932O1WllaWmJ6eprl5WWWlpbWXVBpaSklJSU88sgj1NfXU1RUhNlsTprEvkKhoLa2ltraWqqqqsSKCLhZjOx2u5HL5QSDQWQyGWq1mpqaGqqrq/nKV74ipsilyu1/v/j9foaGhmhra0vKZ7ndbpxOJ8FgELlcjtFo5MiRI/z+7//+XceWU4lOp6OxsZH09PR1edSJ5p6MMxAI4Pf7cTgcOJ1O5ufnuXjxIgsLC6ysrLC4uChunYSEA51OR1NTE1VVVRw+fJjKykpycnKS6hCQSqVizmlLS4tYOLu6uipOIqFQCJ1OJ4Z0mpqaqKuro62t7WPhvPggIpEILpcrKdsz4YgTCoWIRqNiOlxubi7l5eUJ//zNRKVSkZ+fj0ajQalUipVAiV5J78k4LRYLIyMj/MM//ANjY2MsLi6yuLi44WyiVCrZsWMHTzzxBM8//zxlZWUpSR+Dm8aZk5PDpz/9aR544AF++tOf0t7ezltvvbXufQ8++CCtra185jOfIT8//2O3hd2IzaqY+aeIRqMROw7YbDasVisOh4PFxcWEFhDck3GOjIxw5coVrly5wtramliipNVq0Wq17Ny5U6zLPHLkCEVFRezYsYPCwsItsfpkZGRgNBp5/PHH2bNnD5/97GfFB1cikYjezIKCgpRNJJuFTqejoaGBrq6uu1ZnuB8MBgOFhYVUVFSICnUfdx544AHkcjk///nPaW9vJxwOk52djV6vT0ibkHsyTq/Xi9/vRyKRiAYJN5O3jUYjBw4cECsmnn76aYxG45Z6yAVlwPr6+lQPJeFkZmZSU1NDWVkZHo8Hg8GQlAkyMzOT/Px8Ghsb0Wg0rK6uYjKZUi7Fea9IJBLq6+tZW1vjV7/6FePj40ilUg4fPiyWnm32CnpPLQAjkQixWGxdrBL+MYZ460FfLpdv2XS2fwoI+cCCWoJMJhM7pCUa4XMFZQipVCpOjB9H3G43fX19fOc732FgYIBQKMSxY8d47rnneOSRR+5nAdq8FoDCzd0KW9RtPhhhskyFZ1Roevy7Qnp6Omazmeeee47f/OY3TE5O0tfXx+HDh1lbWyM9PX1TF6KP5xS2zTYpQKFQUFBQwDPPPEM0GkWtVvPGG2/gcDjweDyb7hxKWmfrbbb5XSEej+P3+0UhNbVajUqlup/dyYYWvW2c22yTejY0ztT2RNhmm23uyLZxbrPNFmXbOLfZZouybZzbbLNF2TbObbbZomzHOTdgeXkZj8eDw+FAoVCQlpZGSUkJCoXiY1PmtE3qcbvd2Gw24B/lTO+GbePcgNdff5333nuP1157jfz8fCoqKvif//N//s5UqGyTHK5evcp/+A//gVgsRlNTE9/97nfv6u9TZpzhcJhIJEIoFGJhYYHFxUW8Xi/FxcXU1tamJP8yHA7jcrnEXqKCyLXQKFWj0Wwp41xbW8Pv9xOLxbDb7dhsNt54440PrTMsKyujtLSUz33uc/elsXsvCOJqMzMzjI2N0dvbyx/8wR9seXnMuyEWi3Hy5EkuXrzI6OgoDQ0N96QVnFQLEJLlvV6vWNnidrsZHx9namoKp9NJY2Oj2HIg2VvIaDSKz+fDZrOxuLiIRCLB5/OJ/VySWXL1QWMU+m/OzMywtrZGNBrFYrEwPj7Oyy+/LFYM3Ymmpiaampp46KGHRC2fZBmo0B16aGiIrq4uzpw5I1Yu/S4USASDQdxuNxcvXqS7uxu73Y5Go9n6xrm0tITFYuG1114TRbQmJiaIRqOickJDQwO9vb386Z/+adLFjxUKBXq9nvr6ehwOB++9996WK1CenZ1lenqa06dP09HRgdVqJRaLsbq6KrYN/LCHfHBwkLm5OZRKJU8++SRHjhxBrVYnxTiCwSBLS0t87WtfY3V1laysLOx2O7m5uUlXok8EV65c4eTJk/zkJz9hdXUVhUJBfX09dXV1d/2/Em6cTqeTpaUlbty4QV9fH7Ozs/T09LC8vIzP50MqlaJUKpFKpTgcDmw2Gz09PVitVjIyMpLa10MYy601qluNrq4u+vv7OXfuHLOzs7hcLlH28lYhtQ8ytGg0isfj4cqVK9TW1rJz505KS0uTtnLF43ECgYAoouZ2u/H7/b8Txrm6uorFYsHj8SCXyzGZTNTX199T79eEGaegITM/P8/IyAgnT57k7Nmz2Gw28UykVCoxm81iqc3a2hoOh4Pr169js9nIyclJqnEKCnEqlSohle2bwbVr1+jo6ODSpUvi76RS6QeWZ21UexsKhcS270tLS5SUlCRdCT4cDuPxePB4PPj9/qR+9kclFouJR4kPU8uPRCKsrq4yOztLJBIhOzubiooKampq7kl4PCHG6fF46Orq4p133uH06dOsrKywuroqNnqFm6oJZWVlfPe730Uul7O2tsbx48dZWVkRuzolW8RX6A49NTWFxWLZcltagMbGRlwul9jjUiaTkZ2dTWtrK1VVVcTj8XUroNPpZHJyksHBQVwuV6qGfUeSJdV5L8RiMbFn5+nTp/mzP/szduzYsaH6YjAY5N1336Wjo4OJiQmMRiP79+/nq1/9KoWFhff0+ZtunNFolKmpKQYHB+ns7GRmZgafzyeuloLk5N69e2lra6O8vFxspHurtqpQPZ9MNmpFLoxneXkZp9OJ1+tNqgPldiorK8UzJvxjn8nq6mrMZvM6LSS4Kfk5Pz+Pw+FYt/VVKBTk5+eTn5+PwWBIqTMmGo1uSSONxWJcvHhRNNCFhQVyc3M3NM5wOExXVxdjY2P4fD727dtHS0sLZWVl91xwnhDjvH79Op2dnVy8ePF9r6elpWEymXjmmWc4duwYubm5oiJ4qhE8iS6Xa90qE4lEmJqawmg0ih2YU5WM0NraSn19PUeOHAH+sX260KvzdsLhMD6fj87OTlZWVkTjTEtLo6WlhZqaGkpKSlJ2PVKpVJSa3GpEIhFef/11hoeHsdvtzM7OUlhYSH5+/vveGwwG+fWvf83k5CR+v59jx46xa9eu+2o5senGGQ6HOX/+PDdu3BB/l5GRgV6vp7S0lIqKCl544QWqq6tF97nD4WBmZuZ956Jko1QqycnJ4fDhw0gkEl599VWxkdEPf/hDpqen8fv9PPnkkyk7k8pkMtLT09d5sm/XbbqVlZUVent7sdls6xTwBJnQzMzMhHZz+zBisRgWiwWTyURtbW3KxrEREomE9PR08awpxOZvx263Mzk5ic1mw+12f+B774ZNN06pVEpBQQFLS0u4XC5RoDkvL08MftfV1aHVasVEAyE2lOoznrAKVVdX43Q6kUqlYphHaKAqeEdTPc6PYlCC+v7g4CBra2vrVqe0tDRqa2uTHl+USCSi0JfgXFldXd1y52FBLF1oRa9Wq9FoNO/rNBaNRrFarYyMjOD1ekUlvpycnPv2+G+6caanp/PlL3+Z/v5+Ll++zP79+2vv2cEAACAASURBVMnLy8NkMpGdnb3hQ+X3+1ldXU35Qy9w5MgR9Ho93/rWt1I9lHsmFovR2dlJf38/Z86cYXFxcZ1xZmZm8txzzyU9ZCSs8iqVirS0NLGdx1Y52ghcv36dkydPMjw8TDwep7i4mLKyMgoKCsT3CCGsq1ev8qtf/Qq3243RaKSiooIDBw7cdxe1TTdOiUSCXq+ntbWV8vJysrOzxUajd0rJW1paYnR0NCndr+6Gj5tKejgcprOzk8HBQa5du8bw8DDLy8tiTFkgNzeX4uJiFApF0sMnSqVS1NKNxWJia8WtQjwex+Fw0N/fz29+8xtCoRANDQ288MILlJeXrzvOOJ1OfvzjH3P69Gm6u7vJysriiSee4Etf+hK5ubn3fY5PSChFqVRiNBoxGo3rfr+2tkYoFHrfAy+ky0UiERQKhdjtK9WyilvRgyjkIwuB+1vPNaFQiM7OTrq7u7l8+TJ2u13sVSKTyUSl+/LyclGVP9leWqlUikKhQKfTbbkuY+FwmEAgwNDQEKOjo0xPT5OZmUl5eTl79uwhOztbXGBcLhdzc3N0dHRw48YNHA4HtbW11NXVsWfPnk0ZT1LT9377298yMzMjtm+AmyvtwMAAExMTxGIxjEYjlZWVNDY2prxjtEwm23Ir5/LyMjMzM/zmN7/h+vXr2O128bVIJMLVq1c3THxPT09nx44dfOUrX6GlpYXi4uKkpex9XFhaWmJycpKvfe1rYhvIT33qUzz88MPv68x27tw5Ojo6ePPNNwmHw6jVap5//nl27969aeNJmHEKWSl2u525uTmuXr3Ku+++i81mWxfDlEgkYqevjIwMWlpaeOaZZzAYDEnfct2JrfAAC13dXnnlFUZGRhgaGmJlZWVdZk08Hhd3JrePubCwkLq6Oh588EGMRmPS2i5+FHw+X8qKCgRv/PXr1zlz5gxXr14VvfKxWExMNzUYDOzZsweNRkMgEGB4eJj+/n5isRilpaVUVlZy5MiRTW38nBDjFNKynE4nFouFwcFB3n77bXp6enC5XMhkMtHVLDwgEomE3NxcioqK2LNnz5bb8qSaUCiEw+Hg/Pnz9Pb2sri4uOH7BMO8vc17VlYWRqORsrIyMZd5qyCk8KWCQCCAw+Ggq6uLX//612LmlUQiQalUYrPZUCgUKJVK8vLy0Ov1uFwuRkdHsVgswM2+s7t372bnzp2b2hNoU41TCOLfuHGDX/3qV7z88sssLi4Si8XEmJHZbKauro6BgQEWFhbw+/1iH41YLMb09DRnzpzBYDBgNBq3lQdu44O22bc2dr11VYzH4/T09BAMBjl//jzNzc1Jr/i5E/F4nM7OTvR6fUo+/+rVq1y5coXvfOc7BAIBcWLT6XRkZ2czPj7O2NgY586d491330Uul7OysoLdbicQCKDVannmmWf44he/uOk+kk0zzlgsxtLSElarlZdffpm+vj4WFxfFWFpBQQGFhYVkZWWh0+nEuk6h9jAWi+Hz+ZiYmOCdd96htraWyspKiouLU2agt3trb01qTjbp6enk5ubyqU99itraWvr6+sSuXbfen1tXzuHhYRYWFhgbGyMSiYgx21Q7uqRSKcXFxczOzgKsKxlMFl6vF4vFwqlTp7h27RqBQACZTIZWq+Xo0aNiW79Tp06xuLgonvUlEgl+v1/c9vr9fs6fPw/AU089RVZW1qatnptinII8/ezsLP39/bz00ktiycz+/fupra2lvr6enTt3kp6ezvLyMteuXWNsbIylpSWxqiIUCmGxWJiamuLw4cPE43Gys7PJyMhYF4YRupklmtsfYiGJ/E7nukSSlpZGXl4eTz/9NHNzcxQUFNDQ0IDJZNqwoVQsFuONN97g6tWrjI+PJ22cHwWpVIrZbGZsbEz8nTARJuueut1uent7OX36NNevXyc9PR21Wk1eXh6f//znyc7ORiaTMTs7i0wmE0sfbz8ueL1efvOb3zA8PMyBAwfEaMNmcN/GGYvFcDgcvPTSS/z2t7+lu7sbj8dDQUEBNTU1PPvss5SWlqLVavH7/fT29vLtb3+bwcFBFhcXkclk7N27l6amJiwWCxaLheHhYf7mb/6G/Px8WlpaePLJJykpKRFjTEqlksrKyoR/kbd7ay0WCw6Hgxs3bhCPx1OyNczJySE7O5vKykqUSuUH7iq+9KUvYTAY3te5O9VIpVLy8vLEBAhhF2W1WjEajUkJodlsNk6cOEEgEKC2tpbjx4+zZ88eqqqqyM7ORiqVEo/H+dM//VPa29v58Y9/zI0bNzY8G5vNZtra2jCbzZvaf/S+jDMejzM5Ocn4+DgXL15kYmICt9stli81NTVRWFhIeno6LpeL7u5u+vr6GBkZIRAIkJ2dTVtbG3v27KG+vp6KigosFguFhYXMzs4SjUYZGBgQ43NCnqNer6e8vDyhOkNpaWlUV1czPz/P2toagFixMjIyIs6yyUYoCP+w9ovCbmYrSKtsxO1KhoK/Illbbrlcjl6vp6CgAL1ez969e6mqqsJkMonvicfjYpGD1+slFouh0+lobm5Go9GIz19NTQ2VlZWkp6dvqqPtvp7uWCzG5cuXuXDhAqdOnUIqlaLT6fjsZz9LS0sLra2tZGZm4nK5mJiY4JVXXqG3t5fZ2Vny8vIoLy/na1/7mlgZ4fF4ROmSU6dOMTIyQldXF0NDQ+I5z2g0UlVVxQsvvJBQ49RoNBw5coRz586Jxilc86VLl9BoNLS2tibs8++XWCzG8PAwMzMzqR7KhshkMnHnIzT43ShBJVFotVpaW1vZv38/JSUlG060QpjF4XBgtVoJh8OUl5fz1a9+lZKSElEIQKfTkZmZuXUcQisrK5w9e5ZXX32Va9euoVAoePTRR3nwwQc5duwY4XCY/v5+Lly4wOTkJL29vVitVlQqFS+++CJ79+6loqKCxsZGMZlYpVJhNptFaYfp6Wna29u5fPkyKysrAHziE59g9+7dCW/cq1QqKS0t3bB2L1kISnV9fX2o1WrKy8s/klrexMQEfX19/N3f/Z3o7t9KyOVy9u3bx+zsLKWlpdhsNmZmZviHf/gHXnjhhXuS9LhbTCYT/+Jf/AvUavWGz1I4HMbr9fLee+/R29uL1+sVE2SOHj267kiRqObE92ycbrebrq4uLBYLKysr4qwRiUSYnJzE6XQyPT3NpUuXmJubY2xsjNzcXEpLSzlw4ABNTU0UFRWtkyERpDbS0tJEr1c4HEahUOB0OgHEeFKiz5tpaWlUVlaSk5NDRkaGGPKJx+OilKff79/0bsZw0yhdLhcOhwO73U5HRwdFRUUUFhYil8s3fBCEpA+n08mNGze4dOkSo6OjYkK54HS7tag9lWRlZZGTk0NhYSEOhwOlUolOp0ta+ZpCofhAOc61tTXm5uYYGBhgbm4OuVxOU1MTLS0tZGdnJ2WM92ycS0tLvPLKK+KWLxqN0tvby9TUFIFAQPRu3crRo0c5ePAgx48f/0gPidFo5KGHHuKhhx6612HeM1lZWTz55JN0dXVx/fp1cXsYCoV45513UKlUPPnkkxQWFm76diYSiTA0NMTFixc5c+YMHR0dPProo+zfv/+Ohd6BQAC328358+d55513eP3118VzEtxcrYSJb6skIBiNRlpbW5mbm6OiooI//MM/TPiO6KMyNjbG6dOneeutt0TxsW984xtJPcrcs3HK5XJ0Oh1+v59gMEg0GmVpaYnV1VVisRjZ2dns2bOHiooK8vPzqaqqEjVpt8rs/VEwGAwUFRWJZw4hhpjIB9zv9/PSSy8xODjI2NgYwWCQ2dlZfv7zn5OVlbXhWXtmZobZ2VmxsFqIw2m1WsxmM/v376ehoYHGxsZN9ShuFjKZDIVCkfLnQog+COl8wWCQoqIiDh06hNlsTmqJ3T0bZ0ZGBrW1taSlpYn6NFKpVNS0KSkpoaqqirq6OoqKimhqahLd5Kn+Au4GvV5PUVERV69eFQ0z0eOPRCIMDw8zOTnJ8vIycDPhvaur644r59TUFHNzc0xMTBCPx8WAenFxMS0tLRw+fJjKykoMBkNCx34vCDnBHo8HtVqdErV/gUgkwsjICCMjI1gsFrHHya5du9Dr9Uld2e/5LlRWVvKjH/2IK1euMDk5ycjIiChH8vTTT6PValGr1WLN4MfJIG+ltbUVqVTK22+//aFtDhLJzMwMc3NzH/geoaBArVaTmZnJkSNHOHToEF/4whe21Hb2dnw+H3Nzc5w9e5YDBw6Qm5ubsrG4XC6+8Y1vMDMzw+rqKk888QTHjh3jy1/+ctLv3z0bp1QqRa1Ws3PnToqKimhoaBA7cuXm5qJUKtdJUXxc0el0FBQUYDKZROmVRJOens6zzz4rqr719fURDAY3jAHK5XKxzlChUFBSUkJ5eTk1NTU0NzdTXFxMWlralsxRFjLFQqGQGDdOdQ2vQqGgra2NaDSKy+Wirq4Os9mckuf4no1T0LEpLCy8Z13OjwMajYbc3FwqKipQqVSik0tI70rEjiAtLY2HHnoInU6HUqnEarXicDgIBALEYjGxYFmpVJKenk5paSlpaWmoVCqampqor69n9+7dlJeXp1S868PweDzMzs4ikUhQqVTo9fqUj1ehUNDc3Mza2hpOp5Oqqqp10iTJRPIhQd+tVWmcAoTwieBgEe6XXC7f9IyQWxGC8l6vl//3//4fly5d4he/+AVerxetVsv+/fupq6ujvLycBx98UDxGCDE3mUy25Xctp0+f5vvf/z5TU1M0Njby7W9/W5QeTSWhUIhIJEIkEhHP+AneeWw4w2/35/wQhCT7ZEthCjFfmUxGW1sbRqMRs9lMKBQSV0uTyYTBYMBkMpGenp7yVeduKS8v57Of/SwOhwOTybROkTGVfJT0yGSwvXJus03q2XDl3Nr7nm22+SfMtnFus80WZds4t9lmi7JtnNtss0XZNs5tttmibBvnNttsUVIfVNpmG252mltZWeHHP/4x4+PjRKNRvvjFL3Lo0KFUDy1lbBvnNilnZWUFh8OBxWLht7/9LcPDw6Snp/P444+nemgpZds4t0kpsViMH/7wh3R3d3PhwgWxLUdTU1PS2xNuNRJunMvLy2Ln6oGBAQKBACaTCb1eL7ZcKC8vx2w2b8nKiW0Sg8fjYWhoiAsXLnDy5ElR5VCr1VJWVsbzzz9PRUVFqoeZUhJmnNFolGAwyPT0NNPT06KAr9vtpqqqiqKiInQ6HXAzf9VkMpGRkfGxrfvcDARBL0H9PBaLicn20WiUcDgsluHdKaldKHiXy+VbIk/1doRrXFpaoqenh5/+9KcMDw8TCoXIyMjAZDJRXV3NI488siUUGzwej9jXR6vVJlWtIWHf3ujoKB0dHfzd3/0dVqsVj8cjyluOjY2te7gcDgcZGRns3r37Y5e8vVnE43F8Ph99fX1MTk6SlpbG4uIiVqsVv9/P1NQUly9fFhUlNlI0kEqlGAwGduzYwc6dO6mpqUl5feTtOJ1OrFYrf/u3f0t/fz/9/f2Ew2Gqqqo4fvy4OHELws6pJBQK8dd//dcMDAwwODjID3/4Q1paWpKmJpEw48zKyqK0tJRYLEYgEPhAceNr166hUqnQ6XTilvefGj6fj46ODtrb2xkcHEQul+PxeMSGww6Hg5WVFW7cuMHc3NyGVTJCXeT169fp6uqiurpaVLgzm81kZWUlvbrmdgQ5kuHhYWZnZwmHw8DNSpDc3FzKy8spKChIuWEKJXsTExNYLBbsdjsWi4W8vLyPv3EajUYaGhrEguFbub1FXU9PDzMzM2KD0o+jcd56PUKzoFt7kArF6XfC4/Fw+vRpTp8+TW9v7/teF/7HRiLRQoe2W8chk8koKytjx44d7N+/n2PHjq1raZEKhLpYp9PJ6Ogoy8vLSCQS5HK5WGxtNpu3RAc0wTjn5+dZXFwkGAxisVjIz8+noaEhKWNImHEqlUqys7N5/vnn6ejo4MSJE4RCIeRyOQaDgdXVVbHxazAYxOFwcPnyZVH65OOEEKMLhUKsrq5y4cIFLl++LDbqKSkpobq6mj/7sz+7o+ap1+vlzJkzG+oEabVadDod+fn573tNqVRSWFgoNiD2eDwsLy9js9mYnp5mbm6Orq4uUbkhVVX9kUiE+fl5Tpw4wc9//nNR7kUqlfLnf/7n7N69m4ceeijlhdYCwoRRXFws3luXy4Xb7U7eGBL1j4UZsbGxEafTiVarxel0IpVKycjIWNcQRpjtCwoKRCfRVsfv97O2tobdbmd5eZmxsTECgQAul4v+/n4GBweZn58HELf0whZuIzIyMtizZw8lJSW4XC6xJbxEIiEzM1Pc8t+OII7scrnweDxi9yy3243P5yMUCokd4Mxmc2JuxkcgEokwPj7O+Pg4MzMzGI1GcnJyKC8vZ9++fVRWVqZ8y307MpmMyspKVlZWGB0dxe12J7XJb0LdeRKJhP379+P3+8nPz8fn8xGJRFCpVO8LmyiVSh588MGkSPHfL/F4HKfTycDAAOfOnWN4eJhz587h9XpFbVsBiUTC4uIicrn8A/t6Go1GvvnNb7K4uIjb7aasrEy8R+np6eKZ/MPG5fV6eemllxgZGSEUCokTwo0bNz5Q4TzRBAIB2tvbuXHjBmtrazzwwAM8+uijPPfcc+Tm5m5Jz7JUKuWBBx7A6/Vy9uxZlpaWRKnSZJCUO2I2m/nCF77AG2+8wdTUFBMTE+9bRQTt0mQ3Ub0bLl68yLlz55ibm8NqtTI+Po7L5cLn84n9SPV6PQcOHMBkMmEymTh06BBqtVrslHYn5HI5hYWF5ObmEo1G1+n7flDo5FYikQidnZ1ie0VholAqlRw+fJiDBw9u2r24G/7v//2/nD17lsuXLxMKhSguLubLX/4yNTU15OTkbNn4tlQqpbKykv7+fgCuXLmCUqlkZWWFzMzMhEuZJMU4tVotTU1NWK1W1Go1y8vL7zPOWCzG7Owser0+pbqlt3OrY6C3t5eOjg4WFhZYXV1lZWWFvLw89Ho96enppKenk52dzcGDB8nPzycvL499+/Z9JGlKiUQitku4W+LxOMvLyywuLor9a4LBIHBTJbCqqoqqqqqkqyRGo1FWV1dF7/Hs7Cwmk4ni4mLKysowmUxbPnSmUqnEc/Dq6iqLi4ssLCwkRWcoKcZpNBo5cuQIjY2N9Pf3Mzc3x/T0NKurq+J7QqEQr776Kj6fb0s5hPx+P3a7nf/6X/8rly9fpru7G61Wi1arpaamht///d+nsrKS0tJSUapyI8dNIolEIrz77rv8+te/5o033sDr9Yqv7dq1i//4H/8jVVVVST/Tud1u2tvb6e3tZXR0FLi5izp8+DCZmZlbciv7QQiNotrb29FoNAlPkkjK3VlZWWFoaIjZ2VlGRkaYnp7G5/Ote4+whdgKbvRb6e7upqurizNnzojNjT796U+j1+vRaDSYzWY0Gg1qtfojbz83i5mZGWw2G11dXZw9e5bu7m68Xq+4LX7qqac4ePAgRUVFSVeTE5pZDQ8Pr5uElUolGo0GvV6/4WTR1dXFwsKC2CBLpVLx6U9/essYcjweJxwOJ6WPaMKuWGhJt7i4yPT0NJ2dnczOzjI1NcXa2tr7Lk4mk1FUVJTyGGc8HsftdhMIBPB4PGIPkmAwSH5+Pvv37+ef//N/TnZ2dsrkE4Wt9vj4OCMjI5w/f57+/n6mp6eBfzyjNjU1UVNTc8fmR4kkGo3i9XqZm5sT2ydKJBLS09PRarWoVCpxS+tyuURfw8DAAOPj49hsNuBmMkt1dbWYnJLq86mgY/yxNk6Xy8Xc3Bzf/OY3GRkZYWFhgWg0SjQa3fDC5HI5Bw4cSGmycywWIxgM8uabb3Lt2jVOnjzJ008/zZEjRzh8+DA7duygqqoq5TnAPp+PkZER/vN//s+cPXt2w5lcJpNhNBpTlgYnlUoJh8MMDQ2JK2c8HqegoIDm5mbRMOPxOG+++SYrKytIJBLeeOMN0TgFg/7e977HN77xDb7+9a+LXer+KZAw4/R6veLh2el0EgwGP3C2ETyNwAd6NRPJ8vIyly5d4uTJk0xNTZGfny/mqcpkMnQ63ZbokubxeLh27Rp2u33DVu3xeJxgMMiJEyfo7OykpqaG8vJycnJy2LFjB9nZ2QkP9gve7ImJCdxuN1KpVDyrazQaJBIJPp+PlZUVJicnmZycZHp6momJCdbW1pDJZKJQ9vLyMleuXOGVV17h+PHj6yqafpdJmHEKbbuFKosP2wYI7dXz8/PZt29foob1gTgcDs6ePculS5dwuVx84hOfwGw2i86erUIgEGBqaopQKIRCoXjfyimci06fPo1Op6O0tJS9e/dSXl5OLBajvLxc9DAnYhWKx+MsLi4yMzPD/Pw8sVgMpVJJTk4OOp0OlUpFLBZjbW2N6elpLBYLw8PDXLt2TWzIZDQaRQV4Id0vGo3y6KOPolKpto3zfigpKcFkMmGz2Whvb+fEiRMfuHrG43FWVlaSmoFxO7Ozs/yv//W/CIVClJSU8Pjjj1NeXr6lDBNuhkeOHTuG2+0mOzub4eFhPB7PhsUFa2trDAwMMDIyIj74jz32GLt27eK5554jKyuLjIyMTRtbNBrF7/djs9mw2+3AzXOj0WjkiSeeYM+ePRQVFTE3N8fVq1c5efIk77zzDm63G51OR2NjI9XV1Xzuc58jFAphs9n4wz/8Q0KhEG63Oylnva1CwoxT6PXR1tZGWloawWAQv9+/rlbR7/eLgWnhd6m8+QqFAp1Ox+rqKl6vl/b2dtRqNWlpaRQUFKR8OyuQnp5OWVkZx44do7a2lvHxcbxeLz6fD5/Ph9PpZGlpSeyKLTTlgZvx1P7+fjweD5mZmezfv5/6+vpNG5vf72dubo5Lly7R3d1NLBbDbDZTXV3Nww8/TFFREV6vl4sXL9Ld3U1vby9+vx+NRkNlZSVHjx6lsrKS7Oxsrl69Sn9/v9jmsLi4GJ1Ot2XybxNNwtP32traKC4uxmQyidtcQGxTPzw8zNrammiYyTTOW6tG4GZ+644dO8Q8yl/+8peo1WpUKhW5ublbJmCelpaG2WzGbDYTiUSw2Wx4PB7xDDc2NkZfXx+Li4usrq4SiUTEa43H4wwODjIxMYHT6UStVlNdXb1p1+b1epmcnOTMmTNiZk1VVRWHDx/msccew+fzYbfbOXnyJIODgwwMDJCVlUVeXh5tbW184hOfoLi4GIvFwpUrV3j33XdRKBQUFBRQU1ODwWBI6ZZWyHdOxkSdFP+6wWDg8OHD61ZGIT91enqaCxcucO3aNcbHx5mcnMRqtSY83zIWizEyMoJGoyE/Px+FQkFdXR3f//73OXXqFD09Pbz++uv8/d//PefPn+d//+//TWFh4R2rSlKFXC7HZDKtU0zYvXs3Tz31FLt27eL69etcuHCBiYmJdUeGYDDI1atXefvtt4lEInzuc5/blBXJ6/UyOjoqVhwBPPzwwzz77LPI5XJ+9atf8dOf/pT29nY8Hg8ymYy/+Iu/oKWlhebmZpaWlnj33Xf55je/idPpBOBTn/oUjz322JZQR8jIyEhaQkdSjFMmk71vthOStBcWFsSHRkjU/igOpHslEAgwPT3N1NQUIyMjFBYW0traSlFRERkZGZSVldHc3EwkEuFnP/sZKysrKJVK3G63mBK31bh9ElOpVGi1WlpbW9HpdGg0Gt5++21mZ2fXhTUCgQDLy8ui02YziEQiorSH4ATKyclBq9WytraG1WoVPbhCQX5TUxMVFRVkZGQwODjI5cuXmZmZoaioSMwo2rlzZ8q8+LeiUCgwGAxJ8UOkLO0iGo2ytrbGL3/5SwKBAFKpFLlcLjZ9TcS2IRaLsbq6yltvvcVPfvITFhYW2L17N6FQSEwsUKvV7N27F6VSiUQiIRwO4/f7xYqTjwvCkaK1tZWnnnoKt9vNxYsX31fI7Xa7sdvtm2acQvJBLBYjIyODtrY2DAYDkUiE6elp5ufnWV5eJhqNUlFRwbPPPktbWxsZGRnMz8/z2muv0dnZSSwWY//+/TzyyCO88MILW+ZIoVAoktbgN2XGefbsWTo6OsSHQqlUcvToUXbv3p2wSgWHw8Ef/dEfMT4+ztLSEp/5zGfYt28fjz766Lrt0oULF+jo6CAej2M2m6mtrRXjgx83JBIJMpmMkpISSkpK3mecOp2OwsLChIRU7nQ+E37n8XgYHR3lX/7Lf4nf78fj8eD1emlsbOS//bf/RnFxMTk5OVsmdS/Z3PdVC1k109PTqFQq1Go1er3+jiufoL4mOC0E45TJZFRXV1NYWJiwLUM0GsVut4sqDOFwGIfDwY0bN4Cb20OlUsmVK1cYGhpCoVBQVlZGXV0dWq02pWJZsViMaDTKysqKmDublZX1ke/VnZwYGo0Go9G4acapUCjQ6/UolUqi0SgOh4O1tTVcLhcul4tAICB+5263m8nJSYaHh/H5fMTjcerr66mpqeHgwYNkZGQkVe3uo/KxSd/z+XzMzs7y7/7dv6O+vp7GxkY+8YlP3PFBjkQirKys0NHRwZkzZ0TvrVwuZ//+/QktthaSqM+cOcP58+f5P//n/6y7yRqNhoKCAqxWK+FwGJPJxD/7Z/+Mxx9/HJ1Ol9K0sVAohNPp5M0338TtdlNSUsKBAwcoKir6wL+Lx+PEYjHm5uaYnZ1d95pcLic/P5/q6upNW50MBgOHDh3iZz/7GePj47z33nscOXIEnU4npnEKubYzMzPMzMwQj8dRKpVotVoee+wxDhw4sKXKBm9FqDverGPAB3Hf38i5c+fo7Ozk6tWr4mHfbDaLVQdGo5FgMMji4iIrKyssLCxw5swZrl27JgbNa2pqaG5uJi8vL6Fu8oyMDD75yU+i1WrJy8tjenoau90uyonIZDLcbjdms5mCggI+85nPiKGgVM7egvTIiRMn6O7upri4mE9+8pMfqozgdruZm5uju7ubzs5OLBaLnO6CdQAAIABJREFU+JoQnigvL6eqqmrTjhGCPI1Q7xgKhfjFL37BhQsX8Hg82Gy2dasn3AwN5ebm0tzczK5du6ipqdmUsSQCj8dDT08PBw4cSLjH9r6N0263MzU1hcPhIBgM4na76enpIS8vj+zsbAKBAF6vl/HxcaxWKzMzM5w+fVpcnZRKJWVlZezZs+eutmn3glwup7q6Gq/XSzwex2QyMT8/T1ZWlvieeDxOWVkZFRUVPPLII+Tk5Ii5oKkgHo+zsLDA4OAgp0+fFitPDAYDCoVCTA6/9f3BYBCv18vs7Cw3btzgwoULTE1NiaEJQR2hvLycwsLCTZV6lMlkZGRkkJubS15enijnOTg4KH42QGZmJjKZDIVCQVZWFkVFRbS2tmI2m7eEVxb+8bwuTDbhcJhgMMjCwsIHSr1uFvdtnAcPHkSv14vyI0NDQ/zxH/8xWVlZGAwGysvL8Xg8DA4O4vP5iEaj4lY2LS2N0tJSHn/8cT7/+c8nzQhaWlpoamq6Y/mPcD6TyWQpP+9EIhG+//3vc+XKFUZHR0Xjm5+fp6CggKysrHVHCCF++e677/Lj/9/emQfHfdb3/7WHtJdW0kpare77tHXYinX4kuPYjkMchxyEARwgpqGUQjvc04HC9AelYTqllEJLp6WEkkAxkEISH9ix4ziOJct2VrJs3dfqvo/VrnZXe/7+UL8PVuIctqXdTdnXjP+Rpd1nv/t8nuNzvD/PPsv09PSq5A9AZDx9+ctfZuPGjWs6Xr1ez8aNG/nCF75AW1ubUHSfnJwUvxMdHc3+/ftJT08nPT1dCEmXl5eHvCTsjRgMBlJTU8nLy8NisQT1ve/YOJOTk3G73UIM2Ol0curUKWw2Gx6PB5fLxfLysoh9SYagUqlISUnhi1/8IjU1NUEtw5IM772ATCYjMzMTi8Uint3o6Cj/+I//SFZWlkgkl7SApVPK1atXmZqaEqJq0mtFR0eze/duqqurKS0tXZf6WZlMRn5+PvHx8SQkJDA3N7dKnUGhUJCfn09sbKwovI6JiQlLr6y0u0vVSE6nk+7ubnbt2rXu733HT8NgMBAIBMTDViqVNDU1YbPZROKy3+9fFSOUyWTExsaSnZ3Nxz72sbBrGRBOyGQyUUSgUqlwu91MTk7ys5/9jOzsbAwGAzExMcDKkdZutzM1NcX4+Piq11GpVEKF4O6772b37t1kZ2ev24KYmppKamrqmubthgIp/q5Wq1EoFKIi6I1KHuvBmixVBoOBL37xi/j9fhwOB0tLS8Ij19PTw8DAABcvXlx5w/8V6/3MZz7D/v37wya4HK7I5XKqqqrw+/1MTExw7tw55ubmgJUqmrGxsTfdOW9UMJQcNAcOHKCsrIzt27ezceNGTCZTyI/s7xUkvaj+/v5VpZDrzZoYp0wmQ6fT4ff7iYqKYtu2beIIW1JSwvT0NLW1tcDKZJOkGnNyciIT5B2QyWQYDAaKi4s5ePAgOp2O0dFRpqenhbdQUomXyWTExcWJO7PJZEKn06HX68VOWVBQgMFgCMsjZLhiNBrZunUrbW1tuFwuCgsLxWllPVnTb0gul6NWq9m5c+davuwfPfHx8UJLR6rYaG5uFip/r776Kj6fD6VSKcSo5XI51dXVJCcnk5aWRkFBQdgpqr9XSE1NZf/+/TQ2NrK0tMSWLVuCki0me4dMhz+eytb3AFIjILfbzfLysvAmS9k1UrMk6TSiVqtFn04pVzjCrSNltc3MzOD3+9HpdOh0urW8kt30i4kYZ4QIoeemxvnHIWMWIcJ7kIhxRogQpkSMM0KEMCVinBEihCkR44wQIUyJGGeECGFKxDgjRAhTIjlcEcICj8eDzWbj8uXLjIyM0NfXx759+6ioqHhb2Zv/y0SMM0LIkRT7hoeHOXPmDM3NzVy8eBGDwSBUNf4YiRhnhJDT0dHBhQsXeOqpp4Qywkc+8hEqKyv/aHdNiBjnKqRmr9PT09jtdtEzRerlotPpiI+PZ9euXcTExLxnCrYlpHKyS5cuMTw8jNPppLy8nOLiYnQ6XdCNwG63Mz09zbFjx+ju7sZkMlFVVUVmZibFxcXk5OT80fRFuRkR4/xf/H4/c3NzNDY2YjabGRkZEU1cpU7caWlpFBYWUlxcTEZGxnuuysPv9+NyuXjuuec4c+YMMzMzfPaznyU5OVmoKQSTmZkZWlpaePrpp5HJZNx///088cQTFBUVRQrwiRgnbrcbu93Ov/7rv9La2kpjY6Noj56Tk0NSUhIul4vr168zNTXF3Nwcn/70pzl48CCf//znQz38m+J0OrHZbKvauet0OhwOB21tbTQ3N9Pf34/JZCImJobY2NigGmYgEGBsbIzjx4/z/e9/n4SEBGpqavjSl76EwWAIu5aLoSJkxjk1NYXNZhOi1C6XS0iawIrGrMlkory8fF3HMTExQUtLC2azmdHRUbRaLXl5eSQnJ1NQUIBMJsPtdpOUlERPTw99fX10dnaKbmTZ2dkhW+WdTicOh4PR0VHR6g9WJDGlNoA+nw+ZTEZhYSE+n4/u7m5mZ2dxu92ivWEw9Xgl3Vez2Ux3dzeLi4vs2bOHLVu2kJKSEpb3S0nQe3p6moWFBUZGRlZp15pMJoxGIzk5OWv6LENmnFeuXKG9vR2PxyM0b1544QUhOZibm8v999/PD37wg3Udh9ls5pvf/Cbj4+OYTCYeeeQRPvCBD1BYWCgkM71eLwMDA/zHf/wHP/rRj5ienub111/nyJEjPPnkk6LoOdhMT08zODjIL37xC6ampkRDqPn5efFMJe2mxx9/nMTERCwWC7Ozs8hkMqHMHgzJDQnJM/v0008zPT1Nbm4un/rUpygqKgpLw4Q/NGc6f/48zc3N/PznP2d2dlbM1QceeIA9e/bwqU99ak3vyEExTqm9+Pj4OHNzc/T09NDQ0IDFYsHv94vi4RtFwKampnjppZc4fPgwX/nKVygtLV3TMfn9ftra2rh27Rr9/f2UlZVRXV3Nk08+Ke5gEgqFgoyMDGpraxkfH+d3v/sdo6OjnDx5ksceeyxkxjkwMIDZbObEiRO4XC7x/KQVva6uDqfTycDAAOfOnQNWuqwtLi4SExPDvn37KCkpQa/XB80wzGYzDQ0NtLS0UFNTw+HDh8nJyQnLO2YgEKC3t5dr165x8uRJmpubmZ6eZmZmZtWC1tTUhMViYXFxkfr6+jVT5ltT45RU9oaGhvB4POIDzM3N0dvby8jICDMzM3R1ddHb28vMzMybVMykSSIdcxsbG7FarWs5TOAP/UEXFxdxu91kZmaSn59/03YQMpkMjUZDRkYGZWVlHD16dNVx0u/3B/Vo6Ha7xXNsb29ndHQUjUaDRqNBLpcTFxeHwWCgtrYWu92OTqfj+vXrWK1WvF6vaMlYWlq67n1QbyQQCDAyMoLZbMbj8ZCQkEBVVRUxMTEhbXVxM6Tjd0dHB1euXKGhoUHM67i4OHQ6HQqFgoWFBWw2Gzabjddff53c3Nw3CX3fLmv6rSwtLTE5OcmXvvQlhoeHmZ2dBVYmk9PpxOl0ii7Lubm55OXl0d7eftO+E36/H6/Xi9PpXLe+FF6vF41GQ1ZWFvX19WzatOltf1968P/0T//EwsICy8vLOBwOXC5XULstT0xM8PTTT/O73/2Ozs5OfD4fGzZsYMOGDSQmJlJdXc3dd99NQkICNpuNgYEBvvSlL9Hc3IzVakWn02Eymdi9e3dQhKokPB4P7e3tvPzyyxQXF1NSUoLRaAza+98KLpeLubk5fvjDH9LV1cXIyAhxcXFkZmayfft2ampq0Gq1vPjii5jNZgYGBujo6GBsbAyv17smEiZrYpxut5u+vj7Onz9PU1MTV69eFbq1sHLP8Pv9+P1+srKy+PSnP43RaESlUjE2NobD4cBqtfKb3/yG+fl5IUAsKZOvx5FHLpdTVFSEXq+nvLycioqKd2xLIJfLhUaP1+tlcXGR5uZmoqKi2LJly5qP8a2QQiLSrp2YmEhdXR333XcfSUlJGI1GYmNjWVhYoKenh5deeomRkRFcLhfR0dHU1tZSU1ODWq0O2o7ldDp56aWX6OrqwufzcfDgQaqqqoLy3rfDpUuXOHHiBN3d3fj9fioqKnjiiSeEs1CaK9HR0TidTsbHx5mfn2dycpKxsTHS09Pv+ESyJsbp8Xjo6uritdde4/Tp06I5qkR0dLQQnC4qKuLRRx8VvTIWFxeZnJxkeHiY48ePs7i4uDIwpZLY2FjR8XitkclkJCcnYzQaKS8vF4reb4dCoRDizLDSYc1isZCenh5U45TJZKhUKpRKJXK5HJVKRVJSErm5ueTn5yOXy/H7/YyNjdHV1cXFixdFn5SkpCQqKyvZsmVLUNvrLS8vYzabmZiYIDo6mqqqKvLy8oLy3rfD0NAQ586dY2pqSnjuDxw4QEFBgfgdafM5ffo0SqWS5eVlEcZai9Pemhjn4uIiP/3pT7l27RrT09OrBhYdHc2GDRs4dOgQJSUlorMVIDyG165d4/Tp04yOjopdMzk5mfr6ep566ql1a2xzq/G02NhY5HI5eXl5Qs1+fHx8VR+QYGAwGLj33nvp6+tjaWmJoaEhjh07xszMDF/72tdQKpXY7XZ+9KMf0dLSwuuvv45Wq6WwsJBDhw5x8OBBiouLg3rPW15eprGxEZ/Px+bNm8URPFyRVN7lcjkpKSncc889qxpevRG5XI5WqyU2NpaEhIQ1yR67Y+OcmJigt7eX/v5+FhYWVnWqjomJ4X3vex9lZWXU19eTlJT0pmZFUv+J+fl58bdyuZz8/Hzy8vLCSgBZkuaPiooSDz9YjVRvRKVSkZGRQWZmJhkZGVgsFiYnJ+ns7KSnp4fx8XFaWlq4dOkS4+PjeL1eqqqqqKys5O677yYlJSWoqYczMzNYLBYsFgt33XUX9fX1b8pIcjqdTExMoFarUavVYdNF3O/3o1QqiYmJIRAICG93f38/k5OTonteUlIS999/P9u2bROL+J1yx7N+dHSUrq4ukasJK6uOXq8nNTWVRx55hI0bN1JYWPimv5X0QKU8Vp/PJwwgPz+f3NzcdTnS3i4ymQy5XC7+ST8LdnwuOjqalJQUMjIySElJQS6XMzc3J7q8mc1mjh49yuTkJF6vF5VKxZYtW9i2bZtQ3g8mU1NTDAwMMD4+TmpqKnv37kWlUong/vLyMrOzs7S1tREfHy+8odKxPRT4fD58Pp/wvMpkMtGMa3R0lPPnz9PX18fg4CBOp5OsrCw+9KEPkZ2dvWZOtjVpnnvq1CnRzUqpVFJWVkZtbS333nsvu3btuulxwOl0YrVaaWpq4ty5czQ3N+P1esXdqb6+fs3b090pPp9PxGSl+4a00gcTqVuYVqsVXmKbzYbD4eArX/mK6O7m8/nIzs5mx44d7N+/n+Li4qCOU6KlpYVXX32V7OxscnNzSU9PRy6XMzAwQFNTE//+7//O0NAQCwsLREdHk5WVxac+9Sl27tx500U9GCwsLNDb24vP56O3t5ef/OQnPPvssywtLdHT08Py8jLJycn81V/9FeXl5eTm5hIbG7umJ5I7Ns6kpCQyMjJEzZ1Wq2X//v1UVFSwYcMGEQ96I8PDw/T393P8+HHa2tpYWloiNjaWDRs2sHv3bkpLS0lJSbnT4a0pbrebxcVF5ufnxSkhNTV1VYv0G5vXOhwOZmZmiIuLIzExEb1ev6Y7gVarFdcE6RSysLAgjtnp6ekUFxezdetWMjMziYuLW7P3vhUsFgs9PT3cddddZGVliaa/o6OjXLhwAaPRSHJyMjExMczNzeHz+Xj55ZcxGAzodDpSU1NDkj2kVCpRKBQikQNWQizT09MoFAr8fj9paWmkpKSsy/35jo2ztLQUr9dLe3s7UVFRJCYm8olPfAKTyYRer7/p3wQCAdrb2zl//jz/9V//hcfjQaFQkJqayvbt2/nkJz9JUlJS2JVkScY2OTmJ1WpFJpORk5NDRkYGfr9flGRZrVZGRkbE3a+4uJiKigq0Wu2aG6dkcDfee6X+o8XFxdTW1rJv3z4yMzNDVn7V09NDe3s7hw8fpqioCKVSidvtxmKxcOLECT7/+c9TUVFBVVUVly9f5sqVK3zzm9/EZDIRHx+PyWQK+lzQaDTExcVhs9lEs+Ib+5wqlUp0Oh1paWnrVp10x8ZZXl5OYWEhd99998oLKpWkpqa+pRPHZrNhNps5cuQI586dw+PxIJfL0Wg0PPDAA9TV1ZGYmBg2hunz+fB4PMzOzjI0NER3d7dIkwsEAvzkJz/hpZdeoqioiOHhYRYWFrBarczOzoqwUElJCZs2beLLX/7ymjk6/H4/AwMDtLe3rwpbAaJb9N/93d+Rk5ODwWAISavF5eVlRkZGsFqtwrEik8nw+/309vYSCATYsWMHe/bsISsrC41GQ01NDUlJSbS2tjIyMsKRI0fYunVr0OfDww8/zPbt22lra8NqtTI/P88PfvADhoaGAEQ4Ki8vb90WvTs2Tilt7N1OOqvVyssvv0xnZ6fIIEpJSSE7O5uqqiqysrKC5p2VHBKSU8rlcrG0tCRS+mAlhut2uxkdHWViYoKRkRGWl5fFa1gsFnHfGx8fZ3l5Gb1ej16vJy4uDr1eT25uLmlpaWs2wex2Ox0dHXR0dDA4OPgmb3FUVJTIZjGZTGvynreDz+djcXERj8eDTCYjJiZGhK8UCgUmk4m77rqL5ORksftI2UubN2/mtddeo6+vb90yxN6OhIQEUb4mJRfc6JzMz8+nrKxMpEyuB0GPUYyOjvK3f/u3q/IPq6uref/738/BgweD6p2V0goXFxfp6OhgfHyczs5Orl69uir10OVy0dvbe9NJMj4+zsLCgtgl4+Pj2bNnD5WVleTk5BAfH4/BYFjT0MDw8DB///d/z8WLFxkdHQV4U3gqHPD5fMzPz7O8vIxcLic9PV2EGSRPc3V19ZvuwrGxsbzvfe/jwoULIkMnFMhkMvLy8lhYWCAqKmpVXHz79u3cd9996+pNDqpxnjhxggsXLrzpYaelpbFx48ag7Jhut5uWlhaam5sxm82iysBut+N2u8XO6ff7RQs9WFnpb4xpZmRkUFlZyY4dO8jIyCAnJwdY2bWMRiMxMTFoNBoRF12LcdtsNpqamjCbzZw/fx6r1Sq8m0lJSWg0Gi5cuIDb7V4Vcw4HlEoliYmJwrusVqvF83zjiUKpVIrPEw4sLS1hsVhwuVxBjWkHzTgDgQB9fX309vaKn0VFRWEymUhLSwvKpd/r9dLX10dzczOvvfYaV65cwWazYbfbiY6OFvFZo9EoqjukCbK4uIjL5cLtdhMdHU1qaiq1tbXs2rVLJASsJ8vLy8zMzGA2mzGbzUxOTpKQkEBiYiJ33XWXUDOQlBzCJXFDQlKXkL7jt/uuA4GAuNeHQ7WK1+vF4XCIe7NWqxX35/UkKN+g9LAHBwcZHBwUHyo+Pp7Pfe5z1NfXi9jXerKwsMDXvvY1rl27Rl9fH9HR0aSlpbFjxw7S09PJyMigrq6OqKgo1Go1aWlpGAwG7HY7Bw4cEJU2KSkp7Nixg8997nNotdqgOCtmZ2e5dOkSzzzzDD09PQDs37+fHTt28OEPf5i2tjZaW1tRKpXExcWRk5MTEifQjUg7461mUTkcDsxmMzabDY1GE/JjupS/LJfL0ev1bNq0iaSkpHVfANfdOD0eDzMzM3R0dHD16lURLyosLKSkpITt27eTkZERlBXS7XZz/fp1bDYbKSkpfOxjHyMvL4/8/HwR0E9OThYZQGq1WhRkS8n8iYmJPPHEE9TV1aHRaNbdMAOBAAMDA1y8eJGf//znTE1NkZ6ezqOPPsquXbvIy8vD5XLR0tLCqVOn8Hg8aDSasApF3UoWld/vx2q1cvbsWbRaLVu2bAnp55iamqK7u5vz58/j9/tJT0/nnnvuIS8vj/j4+HV973U1TknRzmKxcPnyZQYGBpifnyc+Pp7i4mKqqqooLCwMWk2hz+djfHwcpVJJeno6999/PwUFBaSlpb3pdz0eD3Nzc1y/fp0LFy6wuLgodI12795NYWFhUI6OUoFyR0cHjY2NeDweCgoKeOCBBygtLSUuLo7R0VH6+/tpbW3F6/WKKqBQHwnlcrm4LshkMhwOBzqd7m0LDhYXF5mYmKC9vZ2srCxRZRMKvF4vFotFbCwqlYrMzEyqqqpISUlZ9xredZtdHo8Hh8PBd7/7XV5//XUaGhrwer1kZGTw53/+59x7770UFxe/q1KttUJawdVqNbGxseTn59801ODxeLBYLPz1X/81zc3NDA4OIpfLefDBB/nYxz5GTU1N0JwVfr8fs9lMW1sbi4uLbNq0idraWnbv3k0gEGBhYYFz586JChWfz0d0dDTx8fEh3zlVKhX5+fmkpqYyMDDA8ePH2bZt202L2qUEjmeeeUa0ZPjgBz/IgQMHQnI8l5xqX/nKV+ju7mZubo4nn3ySuro69u/fH5QFY92MU/pw169fp6+vD7fbTWxsLBkZGdTX15ORkRH0jBW5XE5CQoIojj158iR6vR6v10tiYiIajQa9Xk9rays9PT20trbi8XjIzs5mz5497Ny5U2iqBvMeJGUfwR+yf2QymegpcvToUfr7+5HL5VRWVlJTU0N1dXXIBZkVCgWxsbFUVlZit9s5deqUUK0rLi4WYl+zs7NMTk5y7do1rly5gtvt5tChQ1RUVARdtlOiv7+fK1euMDw8jFwup7S0lO3bt1NZWRm0RW9djDMQCIhUt87OToaHh1EoFEI+sKamJiQPXKFQkJmZicViEUrjCoUCh8NBQUEBBoMBk8nE0aNH6enpYWxsjJycHAoLCzl8+DDZ2dkhCeq/0TD9fj92u52enh7MZjMnT55Eq9USHx9PbW0ttbW13HXXXUEf5xuRy+XExMRQUVGBy+XiH/7hH0TVUXx8PB6Ph+npaSE3+vzzz6PT6SgsLOSjH/0oCQkJIVlgfD6fUJCYnZ3FaDSKE0swC8TX3Dj9fj+Dg4M0NzfzyiuvYLPZ0Ol0pKSk8I1vfIPq6uqQed+MRiP/8z//w7lz53j55Zd54YUXsNvt+P1+Lly4ICZOaWkpW7duZe/evZSVlZGZmYlerw9JeCIQCGC1WoXspV6vZ2xsjMOHD9Pe3s7ExAQej4cDBw7wwAMPcM8997xlTnOo2Lp1KyUlJchkMl599VW+/e1v853vfEeUZEle8wMHDvDwww9TUFBAYmJiSO6aHo+HK1eucObMGY4ePYrP5yMrK4v77rtv3R1Ab2TNZ5vX66WlpYUrV67Q3NyMy+UiPj6e8vJycf8IFXK5HKPRSFlZmZD6uFlgOS8vj9TUVDZt2kR6enpIC3+ltDfpjitp1TgcDqE6UVlZKaqAwikvWUKlUgk1QLVaTWZm5ippyejoaLE75eXlkZSUFBLDXF5eZn5+nlOnTtHW1obb7aa6upra2lpKSkqCvouvuXG63W5OnjzJ66+/jtlsRi6Xk5yczN69e8nOzn5bqYdgUVJSQklJCY8++mioh/KOyGQyTCaT6LbV1dUFrOyosbGxJCcnc/DgQbZu3UpRUVHYGaaESqWivr6e+vr6UA/lLbHb7QwPD/Of//mfWK1WlEolH/7wh4WKRLBZU+OcnZ1lbGxMJI7L5XL27dtHVVUVNTU1YWGY7zXkcjm7d+8W5VPPP/88drudnJwc0SHs8ccfJz4+PiyFmd8LSBq1DQ0NnD9/noWFBVJSUigtLWX37t2kp6eHZFxrbpwDAwNMTEzgcDjQarWUlZWxYcMGTCZTpEHNbSCTyUhKSiI/P5/t27eztLTE0tISGRkZFBcXk5WVRWpqashjmu9lPB4PIyMjIuFEo9GQn5/Ptm3bMBqNIcvxXVPj7O7u5syZM1y+fBmlUklKSgr79++ntLT0poH+CO8OSTmvsLCQxx57LNTD+T+H3W7npZde4sSJEzQ1NbFp0yYOHjzIk08+GdINZU2NMzMzk4qKCqEj9OCDD1JcXBzWEogR/rgZHR2lu7ubI0eOYDAYOHToEA888AAlJSUhz01eU+M0GAzk5eVRUVHB5s2bqa2tJSEhIXIXihC2SCp7arWaoqIiSktLqampIT4+PuQJ97J3qBa45eK1QCCA3+8XqXKh/oARIrwTYTBnb/qGa26cESJEuGVuapwRF1+ECGFKxDgjRAhTIsYZIUKYEl5CMxEihDlSNpHkQIqOjl43J1LEOCNEuAV6e3v5wQ9+QG9vLxqNhq9+9atkZ2evS5vKsDBOqbdIS0sLRqORgoIC1Gp1WIZhAoEAMzMzDA0Nce3aNQASExNFbCwS0/2/jcvlwmKx0NfXh1wu5/Tp0+zdu5fExMQ1n69hcee02WwMDQ3x/e9/n6NHj4p2gOGI3++np6eHX//613ziE5/g8OHD/M3f/A0dHR2i5jLCnSMp9t2oAhEOSAXi09PT9Pf38+Mf/5hr166ty3y9pZ1T6qUotSPQaDRER0ffcdL14OAgjY2NmM1m3G43CQkJPProoyQkJIRVQndPTw/Xr1/n17/+NR0dHWKlHBkZ4Xvf+x719fWUlpZy7733CimRCLfO7OwsL774ImfOnGFoaIj3v//97Nixg5qampCOa25ujpGRETo7O1laWsLr9TI8PMzi4iI+n2/Ni/Fv6dXsdjtDQ0MMDg4CUFRURFpa2h2r50kro91uZ3BwkEuXLnHPPfcQExMTch0cWNktp6am6Ovr4/r163R0dDAxMSH+f2lpiY6ODtRqNQ6Hg507dwq191Ag9REdHx9ncXGR2dlZAoEAUVFRJCQkkJWVFbJ2gO+E3+/HZrNx6dIloeGTnJxMYmKi6P0SqgXb5XLhcDiw2+34fD4UCgVxcXHrpil1S7NneHiYI0eO8Oyzz6JQKPjsZz/LgQMHKCoquqNBSCVRSqWSrq4u+vv7eeyxx9Dr9WHRo9PlcnH69Gmhti41T5VwOBwMDAwwMDC1mltDAAAVD0lEQVTAtWvX+MhHPoLJZAqa5OcbkVQSJCW7U6dOicbEe/fu5S//8i+pq6sLydjeieXlZSYmJnjhhReYm5vD5XLxwgsv4PP5kMlkfOADH1h3Scq3wul0iibRsFIttHXrVtLT09clSf62l/aZmRl+8YtfoFarcTqdVFRU3PbqIXW5DpfeHi6XC6fTyejoKKdPn6a5uZnu7m7m5+eZm5vD4/Gg1WpJSEjA7/eLtumhxuPx0NDQQFNTE+fPn8ftdhMTE8NnPvMZFhYWmJmZ4fz58+zfv5+cnJywWPhuxO/3c+nSJZqamsRzVigUxMfHk5WVJXp7Bhvp7nvmzBkuXLggfq7VaoWSZMh3zhtxOp2iI1dMTAz5+fmrGv+8WySF7/Hx8VXGabPZWFpaut3h3RY+nw+Hw8HU1BSzs7O0t7fzyiuv0NTUhNVqxefziTHqdDpycnLw+/0sLCyE1DgDgQBLS0vMz89z+fJloXO7ceNGsrOzqa+vFx3Ujh07htVqFdpJ4XIv9nq9OJ1O0VZieXmZQCCASqUiOzubjIyMkDTRhZW5Pjs7S2trK93d3eLnSqVyTa51b8VtG6c0kf/7v/+bhoYGCgoKyMvLuyXpyEAggNPp5NKlSxw5cgSXywWsGGxDQwNyuZz8/PzbHeItMzs7y8WLF/n9739PR0cHFy9exOv1ivvFjRQWFvLxj39cdOlub28P2jjfiNfrpbGxkddee40f/vCHlJaWcujQIT7+8Y+L43V3d7cI86hUqrDoQXIjNpuNtrY2nnnmGS5duiT8ELGxsXz5y1+msrKS3NzckIytvb2dZ555hueee47Jycmgve8tGWd0dDSJiYmrztcul0towO7bt4/Y2Nh3Levg8/mYnp5mZGQEi8Wyyh19q81v7pTr16/T2trKM888w/DwMHNzcyITRBrrG5HL5SJTRPr/YLv9nU4nly9f5ve//z2NjY3s27ePLVu2sGvXLkwmkzDCoaEhRkZGiI2NJS4uLqz0nKxWK11dXRw5coTx8XHxDFNSUigsLKS0tBSj0Rj0cUlq+6+99hrnz5/HZrOFbwtAtVqNyWRCr9ej1WrF5dhut9PQ0EBeXh4lJSWiI9M7IXnmFhYWmJ+fX2UAKpUqKJXokmezra2NxsZGTp48CfxBwBlubnDSrhPK3Uc6eZjNZlpaWujt7eVDH/oQtbW1VFdXi9/z+XwMDQ0xOjqKRqNBq9WGTe9LQBy5X3nlFebm5oCV55qSkkJBQQEZGRlBd65Jx2zJCdjR0SFaAN4o67me3JJxZmRk8PDDD2Oz2WhsbOSZZ54BVryDDQ0NpKWlodVqefjhh99VpoxCoSAxMZGUlBTS0tIYHR3F5/Mhl8vZuXPnTXtqrDVDQ0M0NTXxve99j/7+fpEvKWmtulwulpeXsdls+P1+5HI5UVFRaDQadDodDocDCI2xOhwOhoaG+Od//mdSUlJ43/vex4c+9KFVOrtOp5OZmRlOnjxJS0tL0Mb2bvH5fHzzm9+kubmZ3t5ecYXQ6XTs2bOH++67j5iYmKA7gjo7O7l8+TLf+c53mJiYwOVyUVRUhN/vX9Vjdj25pU8stcWrrKxkeXmZY8eOiY7QHo+H/v5+Ll26xN69e8Ukfjukld9ut7O4uLhqh1Iqlev6hUit9Zqamvjtb3/L0NAQdrsduVxOUlISGRkZPPLIIwwNDTE0NCTCEQqFgq1bt1JbW0txcTEdHR3i2QSTQCDA2bNnuXz5MgDl5eXs2rXrTcr0o6OjnDhxQoR/NmzYEHYxzomJCSYmJsTJSavVUlVVRVlZGfn5+eviBAoEAni9XhYWFnA4HFitVvGvv78fi8XCwMAAs7OzJCQksHHjRjZt2sT09HR4Gies7Azl5eV4vV6MRiNerxe32w2AxWJBLpdjtVrRaDQ3NU6pmxSsxLTm5uaYm5tjYWHhDj/KuycQCLC8vMz169c5d+4cv/nNb5DJZKIvZ1paGps3b+aLX/wiFy9e5NKlS7zyyisEAgE0Gg179+6lrq6OsrIyhoaGUCgU4v4ZrPG73W5+//vfc+bMGeLi4qiurubAgQOrcpJ9Ph8DAwP84he/oL+/n7i4OGpqakhISAjKON8Jv9+P2+1mZmaGhYUFUd0RGxvLzp07qaioICcnZ93e2+VyMTw8zNTUFBaLhaGhIYaHhzl16hRLS0u43W4UCgW5ubk8/PDD1NbW0tHRwdNPP70uY3ojt7U1abVajEYjVVVVonclrHg7/X4/4+PjaLVadDrdqr+THEDSvW5paYnTp0/T1tZ2hx/j1ujo6MBsNvPd736XsbExANE2Lz8/n8OHD1NVVSV2Q6VSSUFBAfn5+WzatIlDhw6JKoS6ujoUCgWnTp0SjYHXm/n5eS5cuMC1a9dwuVz8v//3/9i8eTN6vV4Ypt/v55VXXuHs2bMiQT8rK4vDhw+HjUxpS0sLR48eZWpqSvxsw4YNVFVV8Wd/9mfr2ptESnb48Y9/zOXLlxkcHMTj8YirS3FxMRs3buSxxx4jKyuLrKwscbIKFrdlnFKPy4yMjFWOBa/XKz600WgkNTWVubk5HA6HUIEfGRnh5ZdfFrtXR0eHcALAipHo9XpUKtW6HWule+bo6KhYWIqKisjNzWXr1q1UVFSQkZEBrLjy09PT2b9/vwiEJyUlic8dExMjHGTBisEtLS1x9epVXC4XsbGxlJSUrGr84/P5cDqdNDY20traitPppKioiMLCQkwmU8hTIn0+H62trVy+fJnm5macTidRUVHExMRQU1NDbW3tm6ICa410ry0rKxPxSliZ23q9XnSXKy8vx2AwoNfrRajvRrxe77qdmG579qvVagoLC9/kkvf5fPT392MymSgoKMBisTAyMkJXVxdOp5PBwUGeffbZt3xdrVZLamqqMND1oLOzk6NHj7K4uIjH40GpVHL33Xezfft2PvjBD676XZPJhFarZdu2bcTExLzpNCA5j6KiooK2qlqtVi5cuIDX6yU9PV2U2Em4XC5mZ2d5/vnn6enpAaCmpoZt27aFxX3T4/Hwq1/9iqtXrwrj1Gg0ZGdn89hjj1FXV7fuYs4qlYq0tDT+9E//VMxL+EPbwri4uHd8Vj6fD5fLtW4VVLdtnImJiTz44IMMDw+jUqloaGgQK/a//Mu/8LOf/YyYmBgcDgdut1t8CCnz462w2WwMDAzQ2NiI1+td0xxQqY241C7C5/Px4IMP8tWvfpXk5OSbts4zGAyihfvb7YzBLG9aXFzk4sWL3H333VRXV6+qDHI4HLz88sv86le/or+/H4PBwL59+zh06BAlJSXrPrZ3YmpqiqGhIc6fP8/g4CDz8/N4vV5SUlJ46KGHyMvLC2oLQ2kHvTHZRaFQ3PS7VqvVxMTEkJSUxOLiIk6nk9dee43c3FxKS0vXfGy3bZxKpRKj0UhxcTGTk5M0NjYCK6vJ8PDwTf/mxiawb4XX68XlcjEyMkJ2dvbtDu9NSOEQ6X4hJa4bjcZVMcE38k5eY7/fvyqtb72x2WzCgZacnExeXh5yuVxUS7S3t9PU1MTrr7+Oz+fDZDKxdetW8vPzQ9L4943Mzs7S29vL+Pg4CwsLLC8vo1arSUxMZOPGjcTFxQU1RU9yQr2buG90dDRarZaYmBicTue6j+2OL3Vbt25Fp9Px05/+9F1v7+8UCwwEAvT29pKZmXmnwxPMzMzQ09PDt771LSYmJkTe7p3GJZeXl3G5XHg8njV5vXeiq6tLuPJLSkqoqqpCJpMxPDzM9evX+frXv874+DhWq5WsrCyqqqr4kz/5E2JiYsKiPWB3dzenTp1icnJSfAepqakUFxezbdu2oDeovRV0Oh3x8fGkpqZitVpRqVRUVlaui0QJBFmmJCkpSdwlpCaparWaa9eurXtu6quvvsrp06cZHx+/6cX+drl69SpXrlyhp6cHu92+7mlmkgNKpVLx4osv0tLSgkKhYGZmhpmZGWZnZ1leXkYmk4ljr06nC3nRutPp5OzZs5w4cYKzZ8+Kk4tSqeSzn/0s1dXVJCQkhLw/ya2gUCgwGAzr5mBbN+OUYoYKhQKVSoVeryc3N1c4VEpLS0lJSUGpVDI9Pb3uxjkyMkJHRwcOh0PkwyYmJt52jqnf78fj8dDZ2UlnZycLCwsoFIp1dwzpdDoMBgNpaWmMj48zNTUlQgBSzDkqKoro6GgRxA9V0beEVCRhNpvp7OxkaGiIQCCAWq0mNjaWmpoaysvLQ+5FvhWkuLhOp1u3BWXdvjXJNS41IT148CD33HMPycnJwMqHc7vddHZ2ijhcMFEqlXzyk59kx44dt/X3drud0dFRfvrTn9La2gpAQkICJpMJlUq1bkfIzMxMoqKi+PrXv87Zs2cZGRnBaDSyefNm0tPT+fa3v01UVBRpaWk89NBDZGVlrcs4bgWr1crg4CDPP/+8MEyFQkFBQQE7d+4kNzc3LLzIt4pCoVjXRl13bJx6vZ60tDRqamrEBT8pKYmcnBxKSkrIyckhOTmZ3NxcDAbDqlVcyqy5GQMDA+Tn5+N0Ot91Iv2tILnMb3W1DgQCzM7OigD6wMAAbrcbnU7HgQMHqK2tJS4ubl1DAXFxcWzbtk0ExtVqNWNjYwwMDGC1Wtm8eTP333//m553sJHS444fP86lS5cYHh7GZrOhVCrZtm0bdXV1HDhwIKzumX6/X1QkSX4E6QQoaWbt2bNHxPSXlpaw2+3Y7XY8Hg+9vb10dXVRUlKCyWS6I7/JHX9zOp0Oo9HIli1bmJmZwWazkZOTQ0VFBXV1deTl5d3yRPX7/YyNjYn74VofFSXPqtvtftcVBlLShNPpZGBggCtXrvDCCy8wOzuLQqEgNTWVnTt3sn379jfFQtcajUZDcXExxcXFwMrneO655+jr68Pv95ORkcHOnTuJiYkJ6V3T6/UyOztLQ0MDJ0+eZHp6WhRQV1VVsW3bNnbu3Bmy8UlIYTCPxyOkL5eWloRvQqFQEB0dTWxsLA6Hg5ycHAwGAzMzM0xMTKDX6wkEArhcLq5cucKrr74qPPghNU61Wk1mZibf+ta3RA2mXC4XIYjbnRyzs7PMz8/j8XjWLHYojS8QCOBwOPjOd77DE088wT333POOf+d0Ojl27Bi//e1vOXv2LDabjeXlZUwmE1u2bOHrX/86eXl5QT+eOZ1OmpubOXbsGMeOHeOjH/0oe/fupaysLOTe2eXlZaGtNDIyIr5HpVLJjh07wiLuKiUSjI2N0djYSEtLC8899xx2u10s3NJ81mq1+Hw+rFariNd/4QtfEBliUvmk0+kUjqI7UQxckzOPXC6/LdElmUyGSqVCp9Oh0+lwOp1iV/P7/UxPT3Pu3Dnq6+tJTk6+ozBFfn4+1dXVdHR04HQ68Xg8OBwOOjo6+OUvfwmsZCelp6ej1+tFjadkzC6Xi6amJlpbW7HZbBgMBvLz8ykvL2fDhg1kZWWFxCtqt9t5/vnn6e/vR6PRsH37dgoLC0NumLDyHUpJKFKYLS4ujtTUVFJTU0N6z5QS381mMxaLhatXr9Lf38/Y2BhTU1Oi0F6n04lFRaVSoVarSU9PX/VaUqqp5JTz+/2UlpbecQ5zSN14MpkMrVaLXq8nISHhTTpCk5OTHDt2jNLSUpKSku5owm3YsIFAIMCLL74ojjGwkgT/ox/9CJlMhtFopLa2lvT0dBwOB88++6zI+vF4PIyPjzM+Po5GoyE/P58HHniAXbt2kZaWFpJK/UAggNVq5Ze//CV+vx+TycSOHTtCMpab4fP5RIqkhNFopKioiJSUlJCqMXg8HhYWFjhz5gwNDQ289NJLwgilO6ak/CFpEGdmZgr/yduhVCqpqakR+dm3S8ib53q9Xjo7O7l+/Trf+MY3mJycFMrpSqUSjUbDr3/9a3bu3HlHrnaPx4Pdbufs2bP89re/5ejRoywtLQnFA+noInlaA4EAdrtdGLJcLic5OZnMzEz27dtHZWXlKn3aUOxUQ0NDtLa28uEPf5g9e/awe/duPvnJT4ZMOvKNtLe38/jjj4s0PYCHHnqIhx9+mEceeQStVhsyJYnR0VGOHz8uFNuljB+VSkVtbS133XUXtbW1VFRUoFarUSgU4nt+N6ETnU53K/Pipg8h5L1SlEolqampAOzcuZO2tjauXr2K1+vF6/WKdDWr1XpHxhkVFYVer2fjxo1MT0/j8/no6uoSGkZS3OqNin/Sl5GYmEhFRYX40rKzs0PqZfT7/bS3t2M2m1EqlWRlZVFeXh42QfylpSXm5uaYmZlheXmZ6OhoUeVRWloqunOFiqioKOLj41Gr1ajVauLi4khMTCQpKYn9+/dTUlJCaWkpOTk5a9LV4HYIuXHCShJ9fHw8f/EXf8Hvfvc7LBYL8/Pz4ug5MTHB2NjYHeeGKpVKiouLSUtLY//+/fzqV7/iwoULjIyMvOXfaLVa4uLi2LJlCw8++CDvf//7V9VNhgKpSPnUqVOcOXOG+Ph4Nm7cSF1dXdgY5/T0NJOTk6LgQaPR8NBDD7F37162bNkS6uGJ5IeCggLsdjvJycls2bKFsrIyHnnkkbB4jmFhnLCyQxUVFXHvvfeiVCp58cUXkcvllJaWUlNTs6YV8VqtlujoaB5//HGqqqqoqKjg+PHjjI2NMTMzI2pRDxw4gNFoJDExkaqqKpKSktDpdCGXlJybm+PcuXM0NzfjcDh46qmnqKqqCqsMG4vFQm9vrxCBk8lkREVFhYWjClaOrykpKXzta19jaWmJqKgodDodWq025BlVEuExiv9Fq9WSlZVFTU0Nc3Nzq3a6tYwdSiVBaWlpIr5ls9kYHx9nenpaGOe2bdswGAzExcVRVFQU8vxUCbvdTktLCwsLC6jVajZt2hTSHiI3Q6PRoNFoxF1NpVKRkJAQNqp/UqQgLy8v1EN5S0LuEIpw61y5coWPfvSjREVFkZubyy9/+cuw7Gfa2trKgw8+iMvlIiEhgZ/85Cfk5uaGRelamBGeDqEIt47P52NpaYnt27ezefNmoqKiws4wAbKzs/m3f/s3vF4vKpWKoqKidc+e+r9ExDjfg6jVarKysti4cSMbNmwIq+PsjcTFxXHfffeFehjvWSLH2vcgkuaqXC4XIaAI72lueuyJGGeECKHnpsYZWXIjRAhT3unOGX5ehggR/kiI7JwRIoQpEeOMECFMiRhnhAhhSsQ4I0QIUyLGGSFCmBIxzggRwpT/D2dEqO3CdwMfAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 데이터셋 분할\n",
        "책에서는 slicing으로 훈련 세트와 테스트 세트를 나누었습니다.\\\n",
        "sklearn.model_seletion.train_test_split()을 사용하여도 같은 결과를 만들 수 있습니다."
      ],
      "metadata": {
        "id": "EMpSAmVCHNMS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y = y.astype(np.uint8)"
      ],
      "metadata": {
        "id": "rhl2-2WPN76Z"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train, X_test, y_train, y_test = X[:60000], X[60000:], y[:60000], y[60000:]"
      ],
      "metadata": {
        "id": "_ZGlRCOqN-eR"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# compare\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "X_tr, X_tt, y_tr, y_tt = train_test_split(X, y, test_size=10000, shuffle=False)"
      ],
      "metadata": {
        "id": "0ULGMsVlOZQj"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(np.sum(X_tt != X_test, axis=None))\n",
        "print(np.sum(y_tt != y_test, axis=None))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6m1ldgTbO7EA",
        "outputId": "393278dc-1ee0-4dd5-d707-c98e1961dd7d"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0\n",
            "0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2. 이진 분류기 훈련\n",
        "이진 분류를 수행하기 위해 0~9로 구성된 label로부터 0(!=5), 1(==5)로 구성된 label을 새로 생성합니다.\\\n",
        "이번에 사용하는 모델은 SGD(Stochastic Gradient Descent)로, 4장에서 상세히 소개됩니다."
      ],
      "metadata": {
        "id": "ysMu94PAO95i"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_train_5 = (y_train == 5) # True if 5 else False\n",
        "y_test_5 = (y_test == 5)"
      ],
      "metadata": {
        "id": "1qyi0AccQLfy"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import SGDClassifier\n",
        "\n",
        "sgd_clf = SGDClassifier(random_state=42)\n",
        "sgd_clf.fit(X_train, y_train_5)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wYGbdBivTpo9",
        "outputId": "0a361313-c678-43f4-f21d-254d31277468"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SGDClassifier(random_state=42)"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sgd_clf.predict([some_digit])       # some_digit == 5"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WYDmpOSnUc6y",
        "outputId": "4048196f-6a1a-4672-a667-8aed620ae258"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ True])"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3. 성능 측정\n",
        "- 교차 검증(cross validation)을 통해 모델의 성능을 검증합니다.\n",
        "- 여러 평가 지표들을 통해 모델을 평가합니다: 오차 행렬(confusion matrix), accuracy, recall, precision, f1-score, roc-auc score"
      ],
      "metadata": {
        "id": "jIH1votNUlTY"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 교차 검증을 사용한 정확도 측정\n",
        "교차 검증은 일정 횟수만큼 훈련 세트와 검증 세트를 바꾸어가며 검증을 진행하는 방법입니다.\\\n",
        "모델의 성능은 초기 가중치 등 무작위한 요소의 영향도 받지만, 훈련 데이터에 따라서도 달라질 수 있습니다."
      ],
      "metadata": {
        "id": "BJ7X5w5bo4s0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import StratifiedKFold\n",
        "from sklearn.base import clone\n",
        "\n",
        "skfolds = StratifiedKFold(n_splits=3, random_state=42, shuffle=True)\n",
        "\n",
        "for train_index, test_index in skfolds.split(X_train, y_train_5):\n",
        "    clone_clf = clone(sgd_clf)\n",
        "    X_train_folds = X_train[train_index]\n",
        "    y_train_folds = y_train_5[train_index]\n",
        "    X_test_fold = X_train[test_index]\n",
        "    y_test_fold = y_train_5[test_index]\n",
        "\n",
        "    clone_clf.fit(X_train_folds, y_train_folds)\n",
        "    y_pred = clone_clf.predict(X_test_fold)\n",
        "    n_correct = sum(y_pred == y_test_fold)\n",
        "    print(n_correct / len(y_pred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7xfY2itLVIAE",
        "outputId": "5f850d47-9b9d-47aa-a456-0a348a1e0ed0"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.9669\n",
            "0.91625\n",
            "0.96785\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import cross_val_score\n",
        "cross_val_score(sgd_clf, X_train, y_train_5, cv=3, scoring=\"accuracy\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KX2jC-FYV4Al",
        "outputId": "b2ee24e7-3f15-460b-b267-1825b140fc0e"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.95035, 0.96035, 0.9604 ])"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "정확도가 90% 이상이라는 의미가 반드시 성능이높다는 의미는 아닙니다.\\\n",
        "사용하고 있는 데이터셋에서 10개 중 9개는 레이블이 0('5'가 아님)이기 때문에, 10개를 모두 0으로 해도 90%의 정확도를 얻을 수 있습니다."
      ],
      "metadata": {
        "id": "gjHJvjmRJwnA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.base import BaseEstimator\n",
        "\n",
        "class Never5Classifier(BaseEstimator):\n",
        "    def fit(self, X, y=None):\n",
        "        return self\n",
        "    def predict(self, X):\n",
        "        return np.zeros((len(X), 1), dtype=bool)    "
      ],
      "metadata": {
        "id": "nsIpWtQMWNHB"
      },
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "never_5_clf = Never5Classifier()\n",
        "cross_val_score(never_5_clf, X_train, y_train_5, cv=3, scoring=\"accuracy\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0w_nr-7nR1sa",
        "outputId": "de38f500-6c30-4b2f-b723-797b6e7d1b55"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.91125, 0.90855, 0.90915])"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 오차 행렬\n",
        "scikit-learn의 confusion matrix는 이진분류 시 다음 형태로 오차행렬을 출력합니다.\n",
        "\n",
        "||Predicted Negative|Predicted Positive|\n",
        "|---|---|---|\n",
        "|Actual Negative|True Negative|False Positive|\n",
        "|Actual Positive|False Negative|True Positive|"
      ],
      "metadata": {
        "id": "uI0E2Qzao-QK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import cross_val_predict\n",
        "\n",
        "y_train_pred = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3)"
      ],
      "metadata": {
        "id": "WRBRLX9jTfBE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "confusion_matrix(y_train_5, y_train_pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Iidit2VqTnLC",
        "outputId": "64b1dda7-cdfd-4199-85a2-d3f0b16df614"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[53892,   687],\n",
              "       [ 1891,  3530]])"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train_perfect_predictions = y_train_5\n",
        "confusion_matrix(y_train_5, y_train_perfect_predictions)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5uvumVMrTwFf",
        "outputId": "a4bb37ed-5a76-46e1-bf98-e35e68e70f73"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[54579,     0],\n",
              "       [    0,  5421]])"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 정밀도와 재현율\n",
        "- 정밀도 precision = TP / (TP + FP) \\\n",
        "- 재현율 recall = TP / (TP + FN)"
      ],
      "metadata": {
        "id": "9TE7Z0o0pDZK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import precision_score, recall_score\n",
        "print(precision_score(y_train_5, y_train_pred))\n",
        "print(recall_score(y_train_5, y_train_pred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ssl501BrWHQO",
        "outputId": "d23d7217-f834-4eb4-d0c9-bd29a5c1a947"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.8370879772350012\n",
            "0.6511713705958311\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import f1_score\n",
        "f1_score(y_train_5, y_train_pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XZhek0ULoc2H",
        "outputId": "bf72e35f-bc28-4abd-84bf-91f76afa97df"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.7325171197343846"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 정밀도/재현율 트레이드오프\n",
        "정밀도와 재현율을 둘 다 개선하는 것은 어렵습니다.\\\n",
        "따라서, 적절한 수준의 정밀도와 재현율을 얻을 수 있는 임계값(threshold)를 정하는 것이 중요합니다.\\\n",
        "scikit-learn에서는 decision_function을 통해 각 샘플이 양성 클래스에 속할 확률값을 얻을 수 있습니다."
      ],
      "metadata": {
        "id": "6iBAdat9ojyZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_scores = sgd_clf.decision_function([some_digit])\n",
        "print(y_scores)\n",
        "threshold = 0\n",
        "y_some_digit_pred = (y_scores > threshold)\n",
        "print(y_some_digit_pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TMBFESE6pIl4",
        "outputId": "997dbdc6-e371-4295-b6f0-73b5974bff53"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[2164.22030239]\n",
            "[ True]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = 8000\n",
        "y_some_digit_pred = (y_scores > threshold)\n",
        "y_some_digit_pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Mxi41FK_pqEs",
        "outputId": "000bc3cb-0c8e-473f-97a9-f0f70ada6a1e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([False])"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_scores = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3, method=\"decision_function\")"
      ],
      "metadata": {
        "id": "KAFdKi1HputO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "precision recall curve를 통해 임계값에 따른 precision, recall의 변화를 확인할 수 있습니다."
      ],
      "metadata": {
        "id": "s6ZXgTzINLRF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import precision_recall_curve\n",
        "\n",
        "precisions, recalls, thresholds = precision_recall_curve(y_train_5, y_scores)"
      ],
      "metadata": {
        "id": "gRogtlC-p6ei"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_precision_recall_vs_threshold(precisions, recalls, thresholds):\n",
        "    plt.plot(thresholds, precisions[:-1], \"b--\", label=\"precision\")\n",
        "    plt.plot(thresholds, recalls[:-1], \"g-\", label=\"recall\")\n",
        "\n",
        "plot_precision_recall_vs_threshold(precisions, recalls, thresholds)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "KJq1hts3qPrT",
        "outputId": "ce4ec950-5d7d-4f29-aeac-a3db8f8da6f8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(precisions, recalls, \"b-\")\n",
        "plt.xlabel(\"precisions\")\n",
        "plt.ylabel(\"recalls\")\n",
        "plt.grid(True)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "id": "KaH1PVdwqpDx",
        "outputId": "c83edad1-f3fc-4557-b3f4-f911f618ad2b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deZhU1bX38e/qbgaZEbQdABkEBCeQFgVjbF6JILmBN4kDeuOsmESicbpxNnGKszcqDjihMYpjDBrUYEJLVJRBFBlEEVAxiogKNsjUrPvHrraalqar6Tp1qqt+n+c5zxmratW2qeXZ++y9zd0REZH8VRB3ACIiEi8lAhGRPKdEICKS55QIRETynBKBiEieK4o7gLpq3769d+7cOe4w0mb16tU0b9487jCygsoiSWWRpLJIqk9ZzJw58wt332FL5xpcIujcuTMzZsyIO4y0KSsro7S0NO4wsoLKIkllkaSySKpPWZjZhzWdU9WQiEieUyIQEclzSgQiInlOiUBEJM8pEYiI5LnIEoGZ3W9mn5vZnBrOm5ndamYLzWy2me0XVSwiIlKzKO8IxgFDt3L+cKB7YhkF3BlhLCIiUoPIEoG7TwG+3MolI4CHPHgdaGNmO0cVz/z5cNJJsHBhVJ8gItIwxdmhbFfg4yr7SxPHPq1+oZmNItw1UFxcTFlZWZ0/7JlndmHcuB6MGwetWm1ghx3WMWTIZ+y55yp2372cxo03bct3qLfy8vJt+j65SGWRpLJIUlkkRVUWDaJnsbuPBcYClJSU+Lb0rCsthZNPhscfh0ceacQHHzTijjt2/+58x46wYgUcfHBY9+oVju28M1RUQP/+UFICjRql6UslqNdkksoiSWWRpLJIiqos4kwEnwAdq+x3SByLzD77hOWqq2DdOpg5E557DtavhylToHv3sDaDrY1iUVAQEkvLltCpE+y0E7jD/vtDjx4heTRpEuU3ERFJnzgTwQRgtJmNBw4AVrr796qFotKkCQwcGJYtcYdvv4VvvgntCtOnw/vvw7vvhjuE1avho49g4kTYsKHmzznwwPBZn34Khx4a7jKaNYODDoKePaP5biIidRFZIjCzR4FSoL2ZLQUuBxoBuPtdwERgGLAQWAOcFFUs28Is/GA3awbFxeGHuyZr1sCCBTB7dkgQM2bAkiXQogWsXQtTp0K7dvDQQ+H85kpp1SrclTRqBNtvD7vvDnvsAbvsEhLRr34VEohZhF9YRPJWZInA3Y+p5bwDZ0T1+ZnUrBn07RuWrXEPP+yzZ8M778DcufDFFx+x/fadmD49VFctWQL//GdYKl17bVj36RMSSefO4U5jr71Cwth7byUJEdl2DaKxOFeYQatW8IMfhAWgrGwRpaWdvnete6hymjQJXn4ZVq2CV1+FRYtCFdWkSZtfv+OO0Lo1DBgQ2iz69IF+/aBrVyjSf2UR2Qr9RGQpM2jcGH7847BUVXln8dprIUm89BJ88AF8/nmoftqS3r3DevnykDT69w/tI4WFMGhQqHoqLIz2O4lIdlIiaIAq7yyGDg3LH/+YPOce7h7eeCO0TcyZA02bhraKKVNg06ZQJTV3LjzwwObvW1gYEkKXLuGpqB12gEMOCY/SqupJJHcpEeQYs1BFdNhhYalJeXnoLzF3bngqavZsWLoU3n4bJk8OS1U77QQ//SnsuSd06xbaJXbdNdrvIiKZoUSQp1q0CMtuu33/nDt88QVMmwYvvBDaIxYsgDtrGA2qQ4eQHFavDk849e4d+lSoqkmkYVAikO8xC9VC1dsnKipCW8ScOaHR+sUXw2OvH3wQ2ioATjll8/fq1Ak2boThw8NTVUcfHe5YRCR7KBFIygoLQ8/pHj3C/nnnbX5+/XqYNQuefTZUM02eHDrdAdx1V1iffnry+hEjQsN1jx6hL8bGjTuy886hh3eBZsoQyRglAkmbxo3hgAPCUlVlVdPkyaGa6amnwg/93/5W/R16c+WVyb199w2d7Lp1g0svDVVOarQWST8lAolcZVXTUUeF5Z57kucqKsLdwGefwf33z2fFil4sWgTLloWGawg9tR97LPmaAQNCu0TTpuEJp8MOC8lH/SVEto3+6UisCgvD4H0tW8KQIcsoLe31vWtefx1eeQX+/e/QI3v9eigrC30iAK64InltUVEY9G/IkFAN1aePEoRIbfRPRLLegQeGpXqbhDu89Rb84x/w5puhk93zz8PHH8O994alUp8+YRjygw8OVUyNG2f2O4hkMyUCabDMah7j6e234cknQye6d94JCePMMze/pnXr0Bdizz1hv/3CfBN9+kD79pmJXyRbKBFITtp337BUWr069LSeNSskhlmzwmOwK1fCvHnwxBObv/6HP4SRI8PQ4d26qU+E5DYlAskLzZvD4MFhqW7t2tAXYv58OPLIcGzKlLBUGjgQjj8+dJTbb7/MxCySKUoEkveaNg3VQ3vuGdod3MMAfjNnhsdd77wzDPD32mvJ1xxzTGi3OP54aNMmvthF0kHddkSqMQuTEQ0bBrfcEu4Y1q0L7Qzt2oVrHn0UzjoL2rYNkwl16wb335+sbhJpSJQIRFLQuHFoc/jii3DHsHIljB0b+i989VUYcuOUU8JgfG3ahGRiFob4/vWvk30iRLKREoHINmjVCk47LfRxcA8d4J5+Ogzt/ctfhmlGGzUK/R3uvDM8jdS0aUgMw4eHYcLd4/4WIoHaCETSYMcdwzDdACeemByp1T2M3HrHHTB9ekgMEMZjqnT++WHcpf79Q/IQyTTdEYhEyCzMK33rreHx1U2bwuOqF18cBtcDuOGGMHVp48bh+qefjjdmyT9KBCIZZBZmfLvqKnjvvTAv9cyZ4SmkSj//OQwaVErLlnDJJWFOiDVr4otZcp8SgUiMiopCv4RHHgnVSC+9lDxXXg5XXw2HHx76Qey0U5iW9Isv4otXcpMSgUgWOfTQkBAmTy6joiI8snr11aHaaNkyuOiiMJKrWej5PG9e3BFLLlAiEMlSBQXhkdWLLgr9GNauDQPpDRkSzj/2WOgEN3Ag3H57aJTWk0iyLZQIRBqIJk1CX4UXXgg/+K++CsceGxqhf/Ob0ChdUACnnhoapUVSpUQg0kANHAh/+UtoSJ40KTm66n33hUHy9tsv2dtZiUG2RolApIHbbrswmN6f/rT500WzZiV7OxcWhv4NixfHFqZkMSUCkRyy3XbJgfO++SZUG51ySmhcfvBB6No1zOA2enQYGkMElAhEclaLFmGE1HvvDVVDkyaFYbQ/+wzGjAmD5XXqBO+/H3ekEjclApE8MXgwTJsGq1YlH0n9+GPo0SP0Z1iwACoq4o5S4qBEIJJnWrZMPpI6blw4VlERnjoqKoJrrgnnJH9EmgjMbKiZLTCzhWZ2wRbOdzKzyWY2y8xmm9mwKOMRkc2dcEJoT3jxRRg6NBy7+OIwUqoZzJgRb3ySGZElAjMrBMYAhwO9gWPMrHe1yy4BHnf3vsBI4I6o4hGRmh12GDz/fJhX4ZproEOHcHz//aFv3zB0tqqNcleUdwT9gYXuvsjd1wPjgRHVrnGgVWK7NfCfCOMRkVp06QIXXhjaDmbPhh/9KAxz8eyzodpo8ODQw1lyi3lEfdLN7AhgqLufmtg/DjjA3UdXuWZn4B9AW6A5MNjdZ27hvUYBowCKi4v7jR8/PpKY41BeXk6LFi3iDiMrqCySsqksyssLufHGnrz88o7fHWvbdj0nnbSY//qvTzGL+vOzpyziVp+yGDRo0Ex3L9niSXePZAGOAO6tsn8ccHu1a84Bzk1sDwDmAQVbe99+/fp5Lpk8eXLcIWQNlUVSNpbFpk3uDzzg3qZNZU+FsMyaFe3nZmNZxKU+ZQHM8Bp+V6OsGvoE6Fhlv0PiWFWnAI8DuPtUoCnQPsKYRGQbmYXeyV99FZaBA8Pxvn2T29IwRZkIpgPdzayLmTUmNAZPqHbNR8ChAGbWi5AIlkcYk4ikQZs2YdC7N98M+1OnhkRx7bXxxiXbJrJE4O4bgdHAi8B8wtNBc83sCjMbnrjsXOA0M3sbeBQ4MXELIyINQN++YSiLiy8O+xdeGBLC88/HG5fUTaT9CNx9orv3cPdu7n514thl7j4hsT3P3Q9y933dvY+7/yPKeEQk/Vq0CFNvrlkDbduGY8OGhbuGqVPjjU1So57FIpIW220HX34Z5mBu0wZWrgxtBw88oAlzsp0SgYik1X77hcbkiRPD/sknhwlzqg6RLdlFiUBEInH44ZvPf9C8ORx3nO4OspESgYhEpnPn8MN/6qlh/+GHw93Bn/4Ua1hSjRKBiETunntg9eowwinAb38bni469li45ZZ4YxMlAhHJkGbNYP58WLEC+vSBdu3g0UfhnHNgwAAoL487wvylRCAiGbX99mE+5S++CAvA66+HeRJ69lSjchyUCEQkNu3ahTaE668P+++9F6qPNOR1ZikRiEjszj8/JITi4jAEdlFRmGtZMkOJQESyxltvwQEHhO3TToPXXos3nnyhRCAiWWOnnUJ7wRNPhP2DDoLf/KYv69fHG1euUyIQkaxzxBHw8sthe86c1jRpAhOqj10saaNEICJZ6Yc/hA0boHnzjQCMGAFjxsQcVI5SIhCRrFVUBM899wrPPBP2R4+GW2+NN6ZcpEQgIllvxAh48MGwfdZZ4U5B0keJQEQahOOPhyFDwnblUBWSHkoEItJgPP10WC9aFMYqkvRQIhCRBqNZM/jss+R+166wcWN88eQKJQIRaVCKi2HGjLC9eDHcd1+88eQCJQIRaXD69YNp08L2L38Jy5fHG09Dp0QgIg1SSQn89Kdhe8cd4cwz442nIVMiEJEGySw0Hj/6aNi/7bbQI1nqTolARBq0kSPhq6/C9lNPwcknxxtPQ6REICINXps2MG5c2H7gAfjmm1jDaXCUCEQkJ5xwAvzkJ2G7Vat4Y2lolAhEJGdUHaH0qqvii6OhUSIQkZwyb15YX3ppmPVMaqdEICI5pVcvaNEibBcUwCefxBtPQ6BEICI55+uvw+xmAPvsE28sDYESgYjknMJCeOWV0Gj85ZcwfnzcEWW3SBOBmQ01swVmttDMLqjhmqPMbJ6ZzTWzR6KMR0TyS+UwFMccAwsWxBtLNossEZhZITAGOBzoDRxjZr2rXdMduBA4yN33BH4bVTwikn969oTjjgvblWv5vijvCPoDC919kbuvB8YDI6pdcxowxt2/AnD3zyOMR0Ty0F13hfX06WovqIl5RM9XmdkRwFB3PzWxfxxwgLuPrnLNM8B7wEFAIfB7d39hC+81ChgFUFxc3G98DlX4lZeX06LyEYc8p7JIUlkkpaMsli9vzFFHDQRg8uSyNEQVj/qUxaBBg2a6e8mWzhXVK6r6KwK6A6VAB2CKme3t7l9XvcjdxwJjAUpKSry0tDTDYUanrKyMXPo+9aGySFJZJKW7LKZNK+V//idtb5dRUf1dRFk19AnQscp+h8SxqpYCE9x9g7svJtwddI8wJhHJU4sXh/Xvfgfr1sUbS7aJMhFMB7qbWRczawyMBCZUu+YZwt0AZtYe6AEsijAmEclTnTsnt5s2jS2MrBRZInD3jcBo4EVgPvC4u881syvMbHjisheBFWY2D5gMnO/uK6KKSUTyW9X5je++O744sk2k/QjcfaK793D3bu5+deLYZe4+IbHt7n6Ou/d2973dPXdagUUk6xQWJsciuvzyeGPJJnVOBGZWYGYa5FVEGqRevcJ62TJYpIpoIMVEYGaPmFkrM2sOzAHmmdn50YYmIhKNc88N627dYOXKeGPJBqneEfR291XA/weeB7oA6qcnIg3SjTcmt9u0gb/+Nb5YskGqiaCRmTUiJIIJ7r4B0EjfItJgbdqUnOz+Zz+LN5a4pZoI7gaWAM0Jnb52A1ZFFZSISNTM4Iknkvsr8vh5xZQSgbvf6u67uvuwxJM+HwKDIo5NRCRyRx8d1u3bQ0VFvLHEZatDTJjZObW8/uY0xiIiknHjx8Njj4Xtfv3grbfijScOtY011DIjUYiIxGjt2tDb+O23YcmSzXsh54OtJgJ3/0OmAhERiUuTJnDZZXDFFbDHHiEx5JPaqoZu3dp5dz8zveGIiMTjggtCIli3DqZOhQED4o4oc2qrGpqZkShERGK23XZwyy1w9tkwaZISwXfc/cFMBSIiEreuXcP68stDVVG+SHWIiR3M7EYzm2hm/6pcog5ORCSTfvKT5PbDD8cXR6al2qHsL4ShpLsAfyB0LpseUUwiIrEwg/feC9u33x5vLJmUaiJo5+73ARvc/WV3Pxn4fxHGJSISi65doVEjeOMNGDcu7mgyI9VEsCGx/tTMfmxmfYHtI4pJRCQ2hYUwcWLYPukk+PjjeOPJhFQTwVVm1ho4FzgPuBc4O7KoRERiNHhw6E8A0KkTfP55vPFErbbHRwFw9+cSmyvRGEMikgfmz4c+fUJv4+Li0MmsSZO4o4pGqk8NPWhmbarstzWz+6MLS0QkftOmJbdPOy2+OKKWatXQPu7+deWOu38F9I0mJBGR7NC4cXJE0j//GcrL440nKqkmggIza1u5Y2bbk2K1kohIQ1ZQAEceGbYvuijeWKKSaiK4CZhqZlea2ZXAa8D10YUlIpI9broprG+7LYxFlGtSnZjmIeBnwLLE8jN3/3OUgYmIZIuOHeH888P2iSfGGkokUr0jgNBvYLW73w4sN7MuEcUkIpJ1rr46rMePB8+xGdtTfWrocuB3wIWJQ42APBqJQ0TyXaNGoU8BhCqiXJLqHcFPgeHAagB3/w+avUxE8swrr4T1H3Jsyq5UE8F6d3fAAcyseXQhiYhkp44dw/rLL+ONI91qTQRmZsBzZnY30MbMTgNeAu6JOjgRkWz1/vtxR5A+tSaCxJ3AkcCTwFNAT+Ayd8+xWjIRkdq1TFSK/+//xhtHOqVaNfQm8LW7n+/u57n7pCiDEhHJVpMnh/Udd8QbRzql2jv4AOC/zexDEg3GAO6+TyRRiYhkqX794o4g/VK9IxgCdCNMRvOTKstWmdlQM1tgZgvN7IKtXPdzM3MzK0kxHhGR2N15Z9wRpEeqPYs/3NKytdeYWSEwBjgc6A0cY2a9t3BdS+As4I26hy8iknmXXBLWuTLcRF16FtdVf2Chuy9y9/XAeGDEFq67ErgOWBthLCIiabP77mF9do5MzxXlCKK7AlUneVtKaGv4jpntB3R097+b2fk1vZGZjQJGARQXF1NWVpb+aGNSXl6eU9+nPlQWSSqLpGwsiw4dAEoBMhpbVGUR21DSZlYA3AycWNu17j4WGAtQUlLipaWlkcaWSWVlZeTS96kPlUWSyiIp28ti4MBSGjfOzGdFVRZRVg19AnSsst8hcaxSS2AvoMzMlgAHAhPUYCwiDUGXxLCbQ4fGG0c6RJkIpgPdzayLmTUGRgITKk+6+0p3b+/und29M/A6MNzdZ0QYk4hIWkyfHtaV/QoassgSgbtvBEYDLwLzgcfdfa6ZXWFmw6P6XBGRTGjXLrn9xBPxxZEOkbYRuPtEYGK1Y5fVcG1plLGIiKTbtdfCBRfAUUc17DkKoqwaEhHJaWedFdZ77hlvHPWlRCAiso0+THSrnTs33jjqS4lARGQb9eyZ3G7Ig9ApEYiI1EPleENnnBFvHPWhRCAiUg+jRsUdQf0pEYiI1ENBlV/RCRNqvi6bKRGIiNTTgw+G9XXXxRvHtlIiEBGppwEDwvq112D9+nhj2RZKBCIi9dS5c3L7xhtjC2ObKRGIiNRTo0Zw0UVhe9q0eGPZFkoEIiJpcNRRYf23v8Ubx7ZQIhARSYN9901uv/12fHFsCyUCEZE0GTIkrPv0iTeOulIiEBFJk+efD+uOHbd+XbZRIhARSZOPPgrrTZvijaOulAhERNKkSZOwbmhzEygRiIikSatWYV1REW8cdaVEICKSJl9+GdbLlsGGDfHGUhdKBCIiabLLLsntG26IL466UiIQEUmTggJ46qmwPWZMvLHUhRKBiEgaNWoU1j/6Ubxx1IUSgYhIGlU2GD/4IEyZEm8sqVIiEBFJo0MOgWHDktsNgRKBiEia/f3vye2G0KdAiUBEJEL33BN3BLVTIhARicCFF4b16afHG0cqlAhERCJwzTVxR5A6JQIRkQg0pIHnlAhERCL27bdxR7B1SgQiIhEoqPLr2qxZfHGkItJEYGZDzWyBmS00swu2cP4cM5tnZrPN7J9mtluU8YiIZFLVUUjXr48vjtpElgjMrBAYAxwO9AaOMbPe1S6bBZS4+z7Ak8D1UcUjIpJpVe8KKucqyEZR3hH0Bxa6+yJ3Xw+MB0ZUvcDdJ7v7msTu60CHCOMREcm48vLkdt++8OGH8cVSE/OIur2Z2RHAUHc/NbF/HHCAu4+u4frbgc/c/aotnBsFjAIoLi7uN378+EhijkN5eTktWrSIO4ysoLJIUlkk5UJZvPDCTlx33R4AFBZu4qWXtm0QovqUxaBBg2a6e8kWT7p7JAtwBHBvlf3jgNtruPYXhDuCJrW9b79+/TyXTJ48Oe4QsobKIkllkZQrZVFR4R4GnHA/4ohte4/6lAUww2v4XY2yaugToGOV/Q6JY5sxs8HAxcBwd18XYTwiIrEpKICbbgrbTz4ZbyzVRZkIpgPdzayLmTUGRgITql5gZn2BuwlJ4PMIYxERid055yS3s2kwusgSgbtvBEYDLwLzgcfdfa6ZXWFmwxOX3QC0AJ4ws7fMbEINbyciklMuvTTuCJKKonxzd58ITKx27LIq24Oj/HwRkWwzfTrsvz9cfTX8/vdQFOmvcGrUs1hEJIO6d09uv/defHFUpUQgIpJBrVvD4ERdSLY8FatEICKSYS+9FNZm8cZRSYlARCTDTjstrDt1ijeOSkoEIiIZdvfdYd2uXbxxVFIiEBHJsG++Ceuvv443jkpKBCIiGbYuMYZCNjw6CkoEIiIZt3BhWF97bbxxVFIiEBHJsD/+MaxPPTXeOCopEYiIZNizz4a1+hGIiEhWUCIQEcmgbBp1tJISgYhIBmXjJPZKBCIiGXTXXWFtBmvWbP3aTFEiEBHJoMrxhdyhrCzWUL6jRCAikkFnngn77hu2f/zj7KgqUiIQEcmwV19Nbm/aFF8clZQIREQyrOqP/513xhdHJSUCEZEMa9kyOTnNY4/FGwsoEYiIZNyGDbBsGRQWwrhxcUejRCAiknEPPwzvvAMVFfDXv8K778YbjxKBiEiGjRiR3L7oIjjmmPhiASUCEZGM23770Jmsa9ewXzkaaVyUCEREMuzLL+Hww2HxYrjvPhg6NN54smR+HBGR/LB4cfJO4Oab4eST440HdEcgIpIxGzfClVcm9w89NL5YqlIiEBGJ2OrVoQqoVy944AE4+GB47z3YZ5+4IwtUNSQiEoE1a+Bf/4JnnoHHH4dvvgljDD3zDAwfnhx8LhsoEYiIpNHGjVBauvl4QieeGOYnHjgwuxJAJSUCEZFtsHo1fPghLFkC8+bB3LkwZ05Yf/tt8robboDzzostzJQoEYiIEAaCKy+HlSvh669h+XL4/PPvL0uXhh//FSs2f/1OO8Fee8Hpp8OBB8JBB0GHDrF8lTqLNBGY2VDgT0AhcK+7X1vtfBPgIaAfsAI42t2XRBmTiOSe5cvh009h7Vr45BM49tiwXd3AgdC0KaxaFX7UV60KA8CtXBnq8GuaT7igAHbYISwdOkBJCey2W3LZYw9o1y7a7xilyBKBmRUCY4AfAUuB6WY2wd3nVbnsFOArd9/dzEYC1wFHRxWTiGSfDz4IP8ju4f/K3ZPLpk0wZ04rli2DkSPr/1mvvfb9Y19/Df36hf+Db9MGWrcOy447hqW4GNq2DckAvl/HX7m/evXWz29pvz7XplOUdwT9gYXuvgjAzMYDI4CqiWAE8PvE9pPA7WZm7jXlZRHJJRMnhlm6tm6/yOOYOTMs2e7669tSWpr+940yEewKfFxlfylwQE3XuPtGM1sJtAO+qHqRmY0CRgEUFxdTli0TfaZBeXl5Tn2f+lBZJOVPWRRyyCE9Wbq0GQUF4f//zMDME2vYtKmCoqICzGD16kIWL26x1Xc0c4qKnEaNNlFU5BQWOkVFYbuoyCkoCEtp6XK2265is9fW9r+g7lZtv/r51F9fl9d+800jpkxpz7fffhvJ30WDaCx297HAWICSkhIvjSIlxqSsrIxc+j71obJIyqeyGDZs6+frXhaWWGrrL7v1hJKNysrWRvJ3EWXP4k+AjlX2OySObfEaMysCWhMajUVEJEOiTATTge5m1sXMGgMjgQnVrpkAnJDYPgL4l9oHREQyK7KqoUSd/2jgRcLjo/e7+1wzuwKY4e4TgPuAP5vZQuBLQrIQEZEMirSNwN0nAhOrHbusyvZa4MgoYxARka3T6KMiInlOiUBEJM8pEYiI5DklAhGRPGcN7WlNM1sOfBh3HGnUnmo9qfOYyiJJZZGkskiqT1ns5u47bOlEg0sEucbMZrh7SdxxZAOVRZLKIkllkRRVWahqSEQkzykRiIjkOSWC+I2NO4AsorJIUlkkqSySIikLtRGIiOQ53RGIiOQ5JQIRkTynRJAhZjbUzBaY2UIzu2AL588xs3lmNtvM/mlmu8URZybUVhZVrvu5mbmZ5eyjg6mUhZkdlfjbmGtmj2Q6xkxI4d9HJzObbGazEv9GapnOpuEys/vN7HMzm1PDeTOzWxNlNdvM6j+Xp7triXghDMP9AdAVaAy8DfSuds0goFli+1fAY3HHHVdZJK5rCUwBXgdK4o47xr+L7sAsoG1if8e4446pHMYCv0ps9waWxB13hOXxQ8JEzXNqOD8MeJ4wDduBwBv1/UzdEWRGf2Chuy9y9/XAeGBE1QvcfbK7r0nsvk6Y0S0X1VoWCVcC1wFrMxlchqVSFqcBY9z9KwB3/zzDMWZCKuXgQKvEdmvgPxmML6PcfQphfpaajAAe8uB1oI2Z7Vyfz1QiyIxdgY+r7C9NHKvJKYSMn4tqLYvErW5Hd/97JgOLQSp/Fz2AHmb2qpm9bmZDMxZd5qRSDr8HfmFmSwlznPwmM6Flpbr+ntSqQUxen0/M7BdACXBI3LHEwcwKgJuBE2MOJVsUEaqHSgl3iVPMbG93/zrWqDLvGGCcu99kZhT6o10AAAPeSURBVAMIMxvu5e6b4g4sF+iOIDM+ATpW2e+QOLYZMxsMXAwMd/d1GYot02ori5bAXkCZmS0h1IFOyNEG41T+LpYCE9x9g7svBt4jJIZckko5nAI8DuDuU4GmhAHY8lFKvyd1oUSQGdOB7mbWxcwaE+ZmnlD1AjPrC9xNSAK5WA9caatl4e4r3b29u3d2986E9pLh7j4jnnAjVevfBfAM4W4AM2tPqCpalMkgMyCVcvgIOBTAzHoREsHyjEaZPSYAxyeeHjoQWOnun9bnDVU1lAHuvtHMRgMvEp6QuN/d55rZFcAMd58A3AC0AJ4wM4CP3H14bEFHJMWyyAsplsWLwGFmNg+oAM539xXxRZ1+KZbDucA9ZnY2oeH4RE88QpNrzOxRQvJvn2gTuRxoBODudxHaSIYBC4E1wEn1/swcLUsREUmRqoZERPKcEoGISJ5TIhARyXNKBCIieU6JQEQkzykRiNSDmZWY2a1bOb+LmT2ZyZhE6kqPj4pUYWaF7l4RdxwimaQ7AskbZtbZzN41s7+Y2Xwze9LMmpnZEjO7zszeBI40s8PMbKqZvWlmT5hZi8Tr9zez18zsbTObZmYtzazUzJ5LnD/EzN5KLLMS5ztXjitvZk3N7AEzeydxflDi+Ilm9rSZvWBm75vZ9YnjhWY2zszmJF5zdkxFJzlOPYsl3/QETnH3V83sfuDXieMr3H2/xDAOTwOD3X21mf0OOMfMrgUeA4529+lm1gr4ttp7nweckXjvFnx/CO0zAHf3vc1sD+AfZtYjca4P0BdYBywws9uAHYFd3X0vADNrk8ZyEPmO7ggk33zs7q8mth8GfpDYfiyxPpAw8cmrZvYWcAKwGyGBfOru0wHcfZW7b6z23q8CN5vZmUCbLZz/QeIzcfd3gQ8JYwcB/DMxztJaYF7iMxcBXc3stsTw06vq+d1FtkiJQPJN9Uaxyv3VibUBk9y9T2Lp7e6npPTG7tcCpwLbERLJHnWIq+posxVAUWIymn2BMuCXwL11eD+RlCkRSL7plBjPHuBY4JVq518HDjKz3QHMrHmi+mYBsLOZ7Z843tLMNqtaNbNu7v6Ou19HGFGzeiL4N/DfiWt7AJ0S77tFiWqqAnd/CriEMH2hSNopEUi+WQCcYWbzgbbAnVVPuvtywqQ4j5rZbGAqsEdiCsWjgdvM7G1gEmEo5Kp+m2jYnQ1s4PuzzN0BFJjZO4SqqBNrmXdiV8K8DG8RqpQurPO3FUmBHh+VvGFmnYHnKhtfRSTQHYGISJ7THYGISJ7THYGISJ5TIhARyXNKBCIieU6JQEQkzykRiIjkuf8DUCBT4i55yYMAAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold_90_precision = thresholds[np.argmax(precisions >= 0.90)]\n",
        "y_train_pred_90 = (y_scores >= threshold_90_precision)\n",
        "print(precision_score(y_train_5, y_train_pred_90))\n",
        "print(recall_score(y_train_5, y_train_pred_90))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Zx2t6UGtt-Xe",
        "outputId": "68289093-e5ae-49c5-dd5e-4269fb0351fd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.9000345901072293\n",
            "0.4799852425751706\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### ROC 곡선\n",
        "ROC 곡선은 FPR에 대한 TPR의 비율입니다.\\\n",
        "FPR은 낮을수록 좋으므로 결과적으로 ROC 곡선 아래의 면적이 넓을수록(=AUC score가 1에 가까울수록) 좋은 모델입니다.\n",
        "- FPR = FP / (FP + FN)\n",
        "- TPR = TP / (TP + FN)"
      ],
      "metadata": {
        "id": "hFGzX8KXwzom"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import roc_curve\n",
        "\n",
        "fpr, tpr, thresholds = roc_curve(y_train_5, y_scores)"
      ],
      "metadata": {
        "id": "s4xSFTgaxQMq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_roc_curve(fpr, tpr, label=None):\n",
        "    plt.plot(fpr, tpr, linewidth=2, label=label)\n",
        "    plt.plot([0, 1], [0, 1], 'k--')\n",
        "    plt.xlabel(\"False Positive Rate\")\n",
        "    plt.ylabel(\"True Positive Rate\")\n",
        "    plt.grid(True)\n",
        "    plt.xlim(0, 1)\n",
        "    plt.ylim(0, 1)\n",
        "plot_roc_curve(fpr, tpr)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "ASSnmE25xVMh",
        "outputId": "dec54a00-bf45-420e-cb1e-0c0676d7ab57"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import roc_auc_score\n",
        "roc_auc_score(y_train_5, y_scores)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Fud1zH_NxsRx",
        "outputId": "e4827aa3-aca6-499b-f10f-a847a2bc0031"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9604938554008616"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "forest_clf = RandomForestClassifier(random_state=42)\n",
        "y_probas_forest = cross_val_predict(forest_clf, X_train, y_train_5, cv=3, method=\"predict_proba\")"
      ],
      "metadata": {
        "id": "hyGBNZB-1Vny"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_scores_forest = y_probas_forest[:, 1]\n",
        "fpr_forest, tpr_forest, thresholds_forest = roc_curve(y_train_5, y_scores_forest)"
      ],
      "metadata": {
        "id": "FZNEQddk1olk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(fpr, tpr, \"b:\", label=\"SGD\")\n",
        "plot_roc_curve(fpr_forest, tpr_forest, \"random forest\")\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "igLM5nSO9nA0",
        "outputId": "e939ea07-ed0f-43ac-e29a-01b4fe9fd9e1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "roc_auc_score(y_train_5, y_scores_forest)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Eu3cC2DW9zNI",
        "outputId": "7fd85732-a5a5-4ac7-ce1f-ba332e81347c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9983436731328145"
            ]
          },
          "metadata": {},
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 4. 다중 분류\n",
        "셋 이상의 클래스를 구분하기 위해서는 다중 분류기가 필요합니다.\\\n",
        "다중 분류기 중 여러 개의 이진 분류기를 사용하는 경우, 분류 방식에 따라 OvR, OvO로 구분합니다.\n",
        "- OvR : One-versus-the-rest. 레이블을 (타겟, 그 외 나머지)로 두고 각 분류기를 훈련시킵니다. N개의 분류기가 필요합니다.\n",
        "- OvO : One-versus-all. 0과 1, 0과 2, ... , 8과 9와 같이 클래스들을 둘씩 조합하여 여러 개의 분류기를 훈련시킵니다. N(N-1)/2개의 분류기가 필요합니다."
      ],
      "metadata": {
        "id": "ftP3wdYm9_gz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.svm import SVC\n",
        "svm_clf = SVC()\n",
        "svm_clf.fit(X_train, y_train)\n",
        "svm_clf.predict([some_digit])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lhLpSe8R_OIm",
        "outputId": "72f7f682-097f-48d5-f1d9-f3df73d52124"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([5], dtype=uint8)"
            ]
          },
          "metadata": {},
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "some_digit_scores = svm_clf.decision_function([some_digit])\n",
        "some_digit_scores"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9Dup-asF_hE9",
        "outputId": "d60eca58-26d7-44fa-ccfb-93a7ed40edb4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 1.72501977,  2.72809088,  7.2510018 ,  8.3076379 , -0.31087254,\n",
              "         9.3132482 ,  1.70975103,  2.76765202,  6.23049537,  4.84771048]])"
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(np.argmax(some_digit_scores))\n",
        "print(svm_clf.classes_)\n",
        "print(svm_clf.classes_[5])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mLolv11aCMMX",
        "outputId": "bbc89adf-237a-4805-e02a-fd62d41c2f6e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "5\n",
            "[0 1 2 3 4 5 6 7 8 9]\n",
            "5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.multiclass import OneVsRestClassifier\n",
        "ovr_clf = OneVsRestClassifier(SVC())\n",
        "ovr_clf.fit(X_train, y_train)\n",
        "ovr_clf.predict([some_digit])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hgdatr9_CbNb",
        "outputId": "d9916c0e-387f-46ee-8bd7-83b748c8dc76"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([5], dtype=uint8)"
            ]
          },
          "metadata": {},
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(ovr_clf.estimators_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lQGDI_y8EXyi",
        "outputId": "7279f527-e59d-4326-bb87-4bd5dd75f35c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "10"
            ]
          },
          "metadata": {},
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sgd_clf.fit(X_train, y_train)\n",
        "sgd_clf.predict([some_digit])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JRNejGLEEZ2D",
        "outputId": "7eb95ac8-f54b-407d-afb3-2b034d5dcc91"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([3], dtype=uint8)"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sgd_clf.decision_function([some_digit])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w-zfF9RFEg3S",
        "outputId": "46eb4024-1b8e-486c-f505-291f50cea684"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[-31893.03095419, -34419.69069632,  -9530.63950739,\n",
              "          1823.73154031, -22320.14822878,  -1385.80478895,\n",
              "        -26188.91070951, -16147.51323997,  -4604.35491274,\n",
              "        -12050.767298  ]])"
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cross_val_score(sgd_clf, X_train, y_train, cv=3, scoring=\"accuracy\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oWQ3j47VElkS",
        "outputId": "8f8286c8-36ac-436c-83f8-a2dd124d8e7f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.87365, 0.85835, 0.8689 ])"
            ]
          },
          "metadata": {},
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "scaler = StandardScaler()\n",
        "X_train_scaled = scaler.fit_transform(X_train.astype(np.float64))\n",
        "cross_val_score(sgd_clf, X_train_scaled, y_train, cv=3, scoring=\"accuracy\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LSoQFEM4EpXF",
        "outputId": "cf52eaf6-025c-4ff4-b22d-e7e2eb5d1c5d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.8983, 0.891 , 0.9018])"
            ]
          },
          "metadata": {},
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 5. 에러 분석\n",
        "선정한 모델의 성능을 높이기 위해서는 하이퍼파라미터 튜닝, Feature Engineering, 훈련 데이터 증강, 에러 분석 등 여러 방법을 사용합니다.\\\n",
        "여기서는 이미지 데이터의 에러 유형을 확인하는 방법을 다룹니다."
      ],
      "metadata": {
        "id": "h1wtCqWhHWCO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_train_pred = cross_val_predict(sgd_clf, X_train_scaled, y_train, cv=3)\n",
        "conf_mx = confusion_matrix(y_train, y_train_pred)\n",
        "conf_mx"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Pg_mlwQ8E3nG",
        "outputId": "4258952c-fd45-4c05-e4e8-f5321bcf95cb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[5577,    0,   22,    5,    8,   43,   36,    6,  225,    1],\n",
              "       [   0, 6400,   37,   24,    4,   44,    4,    7,  212,   10],\n",
              "       [  27,   27, 5220,   92,   73,   27,   67,   36,  378,   11],\n",
              "       [  22,   17,  117, 5227,    2,  203,   27,   40,  403,   73],\n",
              "       [  12,   14,   41,    9, 5182,   12,   34,   27,  347,  164],\n",
              "       [  27,   15,   30,  168,   53, 4444,   75,   14,  535,   60],\n",
              "       [  30,   15,   42,    3,   44,   97, 5552,    3,  131,    1],\n",
              "       [  21,   10,   51,   30,   49,   12,    3, 5684,  195,  210],\n",
              "       [  17,   63,   48,   86,    3,  126,   25,   10, 5429,   44],\n",
              "       [  25,   18,   30,   64,  118,   36,    1,  179,  371, 5107]])"
            ]
          },
          "metadata": {},
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.matshow(conf_mx, cmap=plt.cm.gray)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 275
        },
        "id": "uHk9hie1MDPH",
        "outputId": "f85356a6-7a4a-4c6e-e9a5-9109f0ac8983"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "row_sums = conf_mx.sum(axis=1, keepdims=True)\n",
        "norm_conf_mx = conf_mx / row_sums"
      ],
      "metadata": {
        "id": "BMIYSyNdN9uR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "np.fill_diagonal(norm_conf_mx, 0)\n",
        "plt.matshow(norm_conf_mx, cmap=plt.cm.gray)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 275
        },
        "id": "xC_aBNJLODVR",
        "outputId": "b6824609-67bd-4e1c-8b87-ed13dfdf5e55"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPoAAAECCAYAAADXWsr9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAALyUlEQVR4nO3dzYvd9RXH8c9n7iSTp5rGWJQ8UIOUSihUy6BWoYvYRZ9IRStYUGw32fTBlkJpu/EfELGLUhhss1HaRSpSi7QV2i66MDTGQGPSYrA2iU1sfJgYSjKTmTldzA3EjHp/o7+T372e9wuEzHA9HpJ5+7v35jffcUQIwIfbWNcLAMhH6EABhA4UQOhAAYQOFEDoQAGdhW77C7b/afuI7R91tUdTtrfa/rPtQ7ZfsP1A1zs1Ybtn+3nbv+t6lyZsf9T2Htv/sH3Y9me73mkQ29/vf00ctP0r26u63ulSnYRuuyfpZ5K+KGm7pK/b3t7FLsswJ+kHEbFd0i2SvjUCO0vSA5IOd73EMvxU0u8j4npJn9aQ7257s6TvSpqMiE9J6km6p9utlurqin6TpCMR8VJEzEr6taSvdrRLIxFxIiL29399RotfgJu73eq92d4i6cuSHu16lyZsr5f0OUm/kKSImI2I6W63amRc0mrb45LWSPpPx/ss0VXomyUdu+jj4xryaC5m+1pJN0ra2+0mAz0i6YeSFrpepKFtkk5J2t1/ufGo7bVdL/VeIuIVSQ9JOirphKTTEfHHbrdaijfjlsn2Okm/kfS9iHir633eje2vSPpvRDzX9S7LMC7pM5J+HhE3SvqfpKF+/8b2Bi0+G90maZOktbbv7XarpboK/RVJWy/6eEv/c0PN9gotRv54RDzR9T4D3CZpp+2XtfjSaIftx7pdaaDjko5HxIVnSnu0GP4w+7ykf0XEqYg4L+kJSbd2vNMSXYX+N0mfsL3N9kotvnnx2452acS2tfja8XBEPNz1PoNExI8jYktEXKvF398/RcTQXWkuFhEnJR2z/cn+p26XdKjDlZo4KukW22v6XyO3awjfQBzv4j8aEXO2vy3pD1p8l/KXEfFCF7ssw22S7pP0d9sH+p/7SUQ83eFOH0bfkfR4/wLwkqRvdrzPe4qIvbb3SNqvxb+ZeV7SVLdbLWW+TRX48OPNOKAAQgcKIHSgAEIHCiB0oIDOQ7e9q+sdlmPU9pXY+XIY9n07D13SUP8GvYNR21di58thqPcdhtABJEu5Ycb2yN2Fs3j3YvtG8YakXq/X9QpaWFjQ2Fjz69D8/HzKHhs3bmz0uHPnzmnVquWdN/H666+/n5UGioglX8yd3AI7jJb7h9TUzMxMylwp739O69evT5mbFaMknTlzJmXuzp07U+ZK0u7du9NmX4qn7kABhA4UQOhAAYQOFEDoQAGNQh+1M9gBvN3A0Ef0DHYAF2lyRR+5M9gBvF2T0Ef6DHYALd4Z1//unaG+sR+oqknojc5gj4gp9U+/HMV73YEPsyZP3UfuDHYAbzfwij6iZ7ADuEij1+j9H1LADyoARhR3xgEFEDpQAKEDBRA6UAChAwVwZlzfihUrul5h2c6dO5cyd2FhIWXuW2+9lTJXyjs/b3p6OmWulHMI57udy8cVHSiA0IECCB0ogNCBAggdKIDQgQIIHSiA0IECCB0ogNCBAggdKIDQgQIIHSiA0IECCB0ogNCBAggdKIDQgQIIHSiA0IECCB0ogNCBAlKOex4bG9Pq1aszRqfJOor4+uuvT5krSWfPnk2Z+9prr6XM3bJlS8pcKe/P784770yZK0lPPfVU2uxLcUUHCiB0oABCBwogdKAAQgcKIHSgAEIHChgYuu2ttv9s+5DtF2w/cDkWA9CeJjfMzEn6QUTst/0RSc/ZfiYiDiXvBqAlA6/oEXEiIvb3f31G0mFJm7MXA9CeZb1Gt32tpBsl7c1YBkCOxve6214n6TeSvhcRS24str1L0q7+r1tbEMAH1yh02yu0GPnjEfHEOz0mIqYkTUlSr9eL1jYE8IE1edfdkn4h6XBEPJy/EoC2NXmNfpuk+yTtsH2g/8+XkvcC0KKBT90j4q+SeNENjDDujAMKIHSgAEIHCiB0oABCBwpIOQU2IrSwsND63IyZF9xwww0pcw8cOJAyN9Pdd9+dMvfpp59OmStJK1asSJk7MTGRMleSNm3a1PrMkydPvuPnuaIDBRA6UAChAwUQOlAAoQMFEDpQAKEDBRA6UAChAwUQOlAAoQMFEDpQAKEDBRA6UAChAwUQOlAAoQMFEDpQAKEDBRA6UAChAwUQOlBAynHPkjQ3N9f6zMWf4Jzj6NGjKXN7vV7KXEman59Pmfvkk0+mzF2zZk3KXEk6f/58ytzp6emUuZK0ffv21me+275c0YECCB0ogNCBAggdKIDQgQIIHSiA0IECGoduu2f7edu/y1wIQPuWc0V/QNLhrEUA5GkUuu0tkr4s6dHcdQBkaHpFf0TSDyUtJO4CIMnA0G1/RdJ/I+K5AY/bZXuf7X0R0dqCAD64Jlf02yTttP2ypF9L2mH7sUsfFBFTETEZEZOZ33wCYPkGhh4RP46ILRFxraR7JP0pIu5N3wxAa/h7dKCAZX0/ekT8RdJfUjYBkIYrOlAAoQMFEDpQAKEDBRA6UIAz7mLr9XqRceJn1kmfkjQxMZEyd8eOHSlzJWnv3r0pc0+cOJEy97rrrkuZK0nHjh1LmTszM5MyV5K2bt3a+sxXX31Vs7OzS+5Y44oOFEDoQAGEDhRA6EABhA4UQOhAAYQOFEDoQAGEDhRA6EABhA4UQOhAAYQOFEDoQAGEDhRA6EABhA4UQOhAAYQOFEDoQAGEDhSQdgrs2rVrW5+beSLnunXrUua++eabKXMl6corr0yZe/XVV6fMPXToUMpcScr6Ud0333xzylxJevbZZ1PmRgSnwAIVETpQAKEDBRA6UAChAwUQOlAAoQMFNArd9kdt77H9D9uHbX82ezEA7Rlv+LifSvp9RHzN9kpJ7f9MZABpBoZue72kz0n6hiRFxKyk2dy1ALSpyVP3bZJOSdpt+3nbj9pu//5WAGmahD4u6TOSfh4RN0r6n6QfXfog27ts77O9L+P+eQDvX5PQj0s6HhF7+x/v0WL4bxMRUxExGRGTWd9gAOD9GRh6RJyUdMz2J/uful1S3rchAWhd03fdvyPp8f477i9J+mbeSgDa1ij0iDggaTJ5FwBJuDMOKIDQgQIIHSiA0IECCB0ogNCBAlKOex4bG4uJiYnW587NzbU+84INGzakzD179mzKXElatWpVytzp6emUufPz8ylzJSnrtus77rgjZa4krVy5svWZzzzzjN544w2OewYqInSgAEIHCiB0oABCBwogdKAAQgcKIHSgAEIHCiB0oABCBwogdKAAQgcKIHSgAEIHCiB0oABCBwogdKAAQgcKIHSgAEIHCkg7BTbjhMurrrqq9ZkXzM7OpszNOl1Wkl588cWUuTfddFPK3IMHD6bMlaSZmZmUuZknD2/cuLH1mdPT05qbm+MUWKAiQgcKIHSgAEIHCiB0oABCBwogdKCARqHb/r7tF2wftP0r2zk/xhNAioGh294s6buSJiPiU5J6ku7JXgxAe5o+dR+XtNr2uKQ1kv6TtxKAtg0MPSJekfSQpKOSTkg6HRF/zF4MQHuaPHXfIOmrkrZJ2iRpre173+Fxu2zvs70v4/55AO9fk6fun5f0r4g4FRHnJT0h6dZLHxQRUxExGRGT9pJ76gF0qEnoRyXdYnuNFwu+XdLh3LUAtKnJa/S9kvZI2i/p7/1/Zyp5LwAtGm/yoIh4UNKDybsASMKdcUABhA4UQOhAAYQOFEDoQAGEDhSQctxzr9eLVava/07W+fn51mdekHE8tSRdccUVKXMl6ZprrkmZe+TIkZS5p0+fTpkrSTt37kyZe9ddd6XMlaT7778/ZW5EcNwzUBGhAwUQOlAAoQMFEDpQAKEDBRA6UAChAwUQOlAAoQMFEDpQAKEDBRA6UAChAwUQOlAAoQMFEDpQAKEDBRA6UAChAwUQOlBAyimwtk9J+nfDh18l6bXWl8gzavtK7Hw5DMu+H4+Ij136yZTQl8P2voiY7HSJZRi1fSV2vhyGfV+eugMFEDpQwDCEPtX1Ass0avtK7Hw5DPW+nb9GB5BvGK7oAJIROlAAoQMFEDpQAKEDBfwfaoXCaPcMKr0AAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "'8'의 오답발생 빈도가 높고, '5'와 혼동되는 경우가 많음을 알 수 있습니다."
      ],
      "metadata": {
        "id": "qSg7Dvn0P7uC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "cl_a, cl_b = 3, 5\n",
        "X_aa = X_train[(y_train == cl_a) & (y_train_pred == cl_a)]\n",
        "X_ab = X_train[(y_train == cl_a) & (y_train_pred == cl_b)]\n",
        "X_ba = X_train[(y_train == cl_b) & (y_train_pred == cl_a)]\n",
        "X_bb = X_train[(y_train == cl_b) & (y_train_pred == cl_b)]\n",
        "\n",
        "plt.figure(figsize=(8, 8))\n",
        "plt.subplot(221); plot_digits(X_aa[:25], images_per_row=5)\n",
        "plt.subplot(222); plot_digits(X_ab[:25], images_per_row=5)\n",
        "plt.subplot(223); plot_digits(X_ba[:25], images_per_row=5)\n",
        "plt.subplot(224); plot_digits(X_bb[:25], images_per_row=5)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "id": "Rpa996GzT0mu",
        "outputId": "69e957ba-7ad6-4bb4-8f18-0411f0c0d922"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 4 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 다중 레이블 분류\n",
        "여러 클래스가 있고 각 샘플이 그 중 하나에 해당하는 경우가 다중 클래스 분류라면,\\\n",
        "하나의 샘플이 여러 클래스에 속하는 경우는 다중 레이블 분류라고 합니다.\\\n",
        "여기서는 (1) 7 이상인지(7, 8, 9), (2) 홀수인지(1, 3, 5, 7, 9) 여부로 2개 클래스를 생성하고 다중 레이블 분류를 수행합니다."
      ],
      "metadata": {
        "id": "oS_S4CHyXEnt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "\n",
        "y_train_large = (y_train >= 7)\n",
        "y_train_odd = (y_train % 2 == 1)\n",
        "y_multilabel = np.c_[y_train_large, y_train_odd]\n",
        "\n",
        "knn_clf = KNeighborsClassifier()\n",
        "knn_clf.fit(X_train, y_multilabel)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GRilMi3IZY1U",
        "outputId": "46fe0429-e4fb-4435-9c0d-786e92d4d2d8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "KNeighborsClassifier()"
            ]
          },
          "metadata": {},
          "execution_count": 88
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "knn_clf.predict([some_digit])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "doFSv63YZyLT",
        "outputId": "84420086-f24d-4312-c820-b63f062d13ce"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[False,  True]])"
            ]
          },
          "metadata": {},
          "execution_count": 89
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train_knn_pred = cross_val_predict(knn_clf, X_train, y_multilabel, cv=3)\n",
        "f1_score(y_multilabel, y_train_knn_pred, average=\"macro\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k4prYh-uZ14I",
        "outputId": "9d5a308f-678a-4a7a-dde5-dbde6b4fa243"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.976410265560605"
            ]
          },
          "metadata": {},
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 7. 다중 출력 분류\n",
        "다중 출력 분류는 한 레이블에 값을 여러 개 가질 수 있는 경우입니다. (many-to-many)"
      ],
      "metadata": {
        "id": "DfvqzhhqRnhD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "noise = np.random.randint(0, 100, (len(X_train), 784))\n",
        "X_train_mod = X_train + noise\n",
        "noise = np.random.randint(0, 100, (len(X_test), 784))\n",
        "X_test_mod = X_test + noise\n",
        "y_train_mod = X_train\n",
        "y_test_mod = X_test"
      ],
      "metadata": {
        "id": "SJeP9H36g7rQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "knn_clf.fit(X_train_mod, y_train_mod)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Me9dJNLToRbd",
        "outputId": "eca5c691-3a31-4f6b-d6ef-7da76c1b7d83"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "KNeighborsClassifier()"
            ]
          },
          "metadata": {},
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "some_index = 0\n",
        "clean_digit = knn_clf.predict([X_test_mod[some_index]])\n",
        "plot_digit(clean_digit)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "id": "tZz5opeUoqWG",
        "outputId": "bf1d810c-8e0e-45e2-8b35-b0e74a0bc690"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "OhAgdD4cpq7n"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}