Нейросети_1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Нейросети 1",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ra21eO61q-iI"
      },
      "source": [
        "## Импортируем нужные библиотеки"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ptsy6rKbQU95"
      },
      "source": [
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OYQfIC7VrGhK"
      },
      "source": [
        "## Загружаем данные"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "TZMlhvY_Qkwf",
        "outputId": "2fe368cf-5eac-4ac7-9abc-bc5eaf0065a8"
      },
      "source": [
        "data = pd.DataFrame([[1,1,0],\n",
        "                     [1,0,1],\n",
        "                     [0,1,1],\n",
        "                     [0,0,0]], columns=['p_1','p_2','target'])\n",
        "data.head()"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>p_1</th>\n",
              "      <th>p_2</th>\n",
              "      <th>target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   p_1  p_2  target\n",
              "0    1    1       0\n",
              "1    1    0       1\n",
              "2    0    1       1\n",
              "3    0    0       0"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UCpi3verrKVR"
      },
      "source": [
        "## Функция для отрисовки граници разделения классов"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KzRRCMatgRrr"
      },
      "source": [
        "def plot_ex(df, name_x1, name_x2, fun):\n",
        "    x_min, x_max = df[name_x1].min() - .5, df[name_x1].max() + .5\n",
        "    y_min, y_max = df[name_x2].min() - .5, df[name_x2].max() + .5\n",
        "\n",
        "    h = .02\n",
        "    xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
        "\n",
        "    Z = fun(np.c_[xx.ravel(),yy.ravel()])\n",
        "    Z = Z.reshape(xx.shape)\n",
        "    plt.figure(1, figsize=(7,7))\n",
        "    plt.pcolormesh(xx,yy,Z,cmap=plt.cm.Pastel1)\n",
        "\n",
        "    sns.scatterplot(x=df[name_x1], y=df[name_x2], hue=df['target'], palette=['#00aa00','#0000ff'])\n",
        "    plt.xlabel(name_x1)\n",
        "    plt.ylabel(name_x2)\n",
        "\n",
        "    plt.xlim(xx.min(), xx.max())\n",
        "    plt.ylim(yy.min(), yy.max())\n",
        "    plt.xticks()\n",
        "    plt.show()"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mR8dHXxEtQo4"
      },
      "source": [
        "## Подберем параметры модели"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M5iGSbl1rf7M"
      },
      "source": [
        "![image.png](data:image/png;base64,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)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L8SmWLQGulVA"
      },
      "source": [
        "### Модель\n",
        "Модель состоит извходного слоя с двумя входами, так как у нас всего 2 признака, из промежуточного слоя из трех нейронов и выходного слоя с одним нейроном.\n",
        "\n",
        "Для промежуточного слоя значения определяются формулами\n",
        "\n",
        "$f_3 = w_{13} x_1 + w_{23} x_2 + b_3$\n",
        "\n",
        "$f_4 = w_{14} x_1 + w_{24} x_2 + b_4$\n",
        "\n",
        "$f_5 = w_{15} x_1 + w_{25} x_2 + b_5$\n",
        "\n",
        "Функция активация на промежуточных нейронах пороговая\n",
        "\n",
        "\\begin{align}\n",
        "    porog(x) = \\left\\{\n",
        "      \\begin{array}{cl}\n",
        "        0 & x \\le 0 \\\\\n",
        "        1 & x > 0.\n",
        "      \\end{array}\n",
        "    \\right.\n",
        "\\end{align}\n",
        "Выходой нейрон \n",
        "\n",
        "$f_6 = w_{36} f_3 + w_{46} f_4 + w_{56} f_5 + b_6$\n",
        "\n",
        "Функция активация на выходе - сигмоида\n",
        "\n",
        "$sigmoida(x) = \\frac{1}{1+e^{-x}}$\n",
        "\n",
        "Порог для принятия решения о том, что объект относится к 1 классу равен 0.5\n",
        "\n",
        "Итог:\n",
        "\n",
        "$f_3 = porog(w_{13} x_1 + w_{23} x_2 + b_3)$\n",
        "\n",
        "$f_4 = porog(w_{14} x_1 + w_{24} x_2 + b_4)$\n",
        "\n",
        "$f_5 = porog(w_{15} x_1 + w_{25} x_2 + b_5)$\n",
        "\n",
        "$f(x_1,x_2) = sigmoida(w_{36} f_3 + w_{46} f_4 + w_{56} f_5 + b_6)$\n",
        "\n",
        "Задача в том, чтобы подобрать такие \n",
        "\n",
        "$w_{13}$, $w_{14}$, $w_{15}$, $w_{23}$, $w_{24}$, $w_{25}$, $w_{36}$, $w_{46}$, $w_{56}$, $b_3$, $b_4$, $b_5$ \n",
        "\n",
        "что бы модель давала верный прогноз по всем случаям"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 618
        },
        "id": "ABOcMLTwgYFC",
        "outputId": "e95da0ae-99b4-4af7-9c7b-c77a0d4deb86"
      },
      "source": [
        "# Подбираем значения параметров модели\n",
        "# =====================\n",
        "w13 = 1\n",
        "w14 = 1\n",
        "w15 = 1\n",
        "w23 = 1\n",
        "w24 = 1\n",
        "w25 = 1\n",
        "b3 = 1\n",
        "b4 = 1\n",
        "b5 = 1\n",
        "w36 = 1\n",
        "w46 = 1\n",
        "w56 = 1\n",
        "# =====================\n",
        "\n",
        "# функция описывающая работу нашей модели \n",
        "# здесь х это массив пар значений признаков\n",
        "# вида [[0.1, 0.2]\n",
        "#      [1.3, 3.1]\n",
        "#      [2.1, 1.2]\n",
        "#      ...       ]\n",
        "\n",
        "def sigmoida(x):\n",
        "    return 1/(1+np.exp(-x))\n",
        "\n",
        "def porog(x):\n",
        "    return (x>0)*1\n",
        "\n",
        "def f(x): \n",
        "    f3 = porog(w13*x[:,0]+w23*x[:,1]+b3)\n",
        "    f4 = porog(w14*x[:,0]+w24*x[:,1]+b4)\n",
        "    f5 = porog(w15*x[:,0]+w25*x[:,1]+b5)\n",
        "    f6 = w36*f3+w46*f4+w56*f5 \n",
        "    res = sigmoida(f6)\n",
        "    return  (res>0.5)*1\n",
        "\n",
        "# Сделаем прогноз \n",
        "data['pred'] = f(data[['p_1','p_2']].to_numpy())\n",
        "\n",
        "# Нарисуем границу разделения классов и выведем результат предсказания\n",
        "plot_ex(data, 'p_1', 'p_2', f)\n",
        "print(f\"Доля верных ответов: {sum(data['target'] == data['pred'])/data.shape[0]}\")\n",
        "data"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 504x504 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Доля верных ответов: 0.5\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>p_1</th>\n",
              "      <th>p_2</th>\n",
              "      <th>target</th>\n",
              "      <th>pred</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   p_1  p_2  target  pred\n",
              "0    1    1       0     1\n",
              "1    1    0       1     1\n",
              "2    0    1       1     1\n",
              "3    0    0       0     1"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eKislGYQ1VN1"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}