8_神经网络.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "6d100c08",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "plt.rcParams['font.sans-serif'] = ['SimHei']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2bbf566e",
   "metadata": {},
   "source": [
    "## 编写神经网络反向传播算法"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9295252a",
   "metadata": {},
   "source": [
    "### 分类（损失函数为交叉熵）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "a2d51906",
   "metadata": {},
   "outputs": [],
   "source": [
    "class NeuralNetwork_clf:\n",
    "    def __init__(self, layer_sizes, learning_rate=0.01):\n",
    "        self.layer_sizes = layer_sizes\n",
    "        self.learning_rate = learning_rate\n",
    "        self.num_layers = len(layer_sizes)\n",
    "\n",
    "        # 初始化权重和偏置\n",
    "        self.weights = [np.random.rand(layer_sizes[i], layer_sizes[i+1]) for i in range(self.num_layers - 1)]\n",
    "        self.biases = [np.zeros((1, size)) for size in layer_sizes[1:]]\n",
    "\n",
    "    def sigmoid(self, x):\n",
    "        return 1 / (1 + np.exp(-x))\n",
    "\n",
    "    def sigmoid_derivative(self, x):\n",
    "        return x * (1 - x)\n",
    "\n",
    "    def cross_entropy_loss(self, y_true, y_pred):\n",
    "        epsilon = 1e-15  # 避免对数中出现无穷大值\n",
    "        y_pred = np.clip(y_pred, epsilon, 1 - epsilon)\n",
    "        return -np.sum(y_true * np.log(y_pred) + (1 - y_true) * np.log(1 - y_pred))\n",
    "\n",
    "    def softmax(self, x):\n",
    "        exp_values = np.exp(x - np.max(x, axis=1, keepdims=True))\n",
    "        return exp_values / np.sum(exp_values, axis=1, keepdims=True)\n",
    "\n",
    "    def forward(self, X):\n",
    "        self.layer_outputs = [X]\n",
    "        for i in range(self.num_layers - 1):\n",
    "            layer_input = np.dot(self.layer_outputs[-1], self.weights[i]) + self.biases[i]\n",
    "            if i == self.num_layers - 2:\n",
    "                layer_output = self.softmax(layer_input)\n",
    "            else:\n",
    "                layer_output = self.sigmoid(layer_input)\n",
    "            self.layer_outputs.append(layer_output)\n",
    "        return self.layer_outputs[-1]\n",
    "\n",
    "    def backward(self, X, y_true):\n",
    "        # 反向传播\n",
    "        output_error = self.layer_outputs[-1] - y_true\n",
    "\n",
    "        self.weight_gradients = []\n",
    "        self.bias_gradients = []\n",
    "\n",
    "        for i in range(self.num_layers - 2, -1, -1):\n",
    "            # 计算梯度\n",
    "            weight_gradient = np.dot(self.layer_outputs[i].T, output_error)\n",
    "            bias_gradient = np.sum(output_error, axis=0, keepdims=True)\n",
    "\n",
    "            # 保存梯度\n",
    "            self.weight_gradients.insert(0, weight_gradient)\n",
    "            self.bias_gradients.insert(0, bias_gradient)\n",
    "\n",
    "            # 更新output_error用于下一层\n",
    "            output_error = output_error.dot(self.weights[i].T) * self.sigmoid_derivative(self.layer_outputs[i])\n",
    "\n",
    "    def update_weights(self):\n",
    "        # 更新权重和偏置\n",
    "        for i in range(self.num_layers - 1):\n",
    "            self.weights[i] -= self.weight_gradients[i] * self.learning_rate\n",
    "            self.biases[i] -= np.sum(self.bias_gradients[i], axis=0, keepdims=True) * self.learning_rate\n",
    "\n",
    "    def train(self, X, y_true, epochs=1000):\n",
    "        losses = []\n",
    "\n",
    "        for epoch in range(epochs):\n",
    "            # 前向传播\n",
    "            self.forward(X)\n",
    "\n",
    "            # 计算损失\n",
    "            loss = self.cross_entropy_loss(y_true, self.layer_outputs[-1])\n",
    "\n",
    "            # 反向传播\n",
    "            self.backward(X, y_true)\n",
    "\n",
    "            # 更新权重和偏置\n",
    "            self.update_weights()\n",
    "\n",
    "            # 记录损失\n",
    "            losses.append(loss)\n",
    "\n",
    "        # 可视化损失\n",
    "        plt.plot(range(epochs), losses)\n",
    "        plt.xlabel('训练次数')\n",
    "        plt.ylabel('损失函数值')\n",
    "        plt.title('训练过程中损失函数变化')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3bebc7b5",
   "metadata": {},
   "source": [
    "### 回归（损失函数为均方差）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "a3fbd586",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "class NeuralNetwork:\n",
    "    def __init__(self, layer_sizes, learning_rate=0.01):\n",
    "        self.layer_sizes = layer_sizes  # # 一个包含每个层维度的列表\n",
    "        self.learning_rate = learning_rate\n",
    "        self.num_layers = len(layer_sizes)\n",
    "        \n",
    "        np.random.seed(1017)\n",
    "\n",
    "        # 初始化权重w和常数项b\n",
    "        self.weights = [np.random.rand(layer_sizes[i], layer_sizes[i+1]) for i in range(self.num_layers - 1)]\n",
    "        self.biases = [np.zeros((1, size)) for size in layer_sizes[1:]]\n",
    "\n",
    "    def sigmoid(self, x):  # sigmoid函数\n",
    "        return 1 / (1 + np.exp(-x))\n",
    "\n",
    "    def sigmoid_derivative(self, x):  # sigmoid函数的导数\n",
    "        return x * (1 - x)\n",
    "\n",
    "    def mean_squared_error(self, y_true, y_pred):  # 均方误差\n",
    "        return np.mean((y_true - y_pred) ** 2)\n",
    "\n",
    "    def forward(self, X):  # 前向传播\n",
    "        self.layer_outputs = [X]\n",
    "        for i in range(self.num_layers - 1):\n",
    "            layer_input = np.dot(self.layer_outputs[-1], self.weights[i]) + self.biases[i]\n",
    "            layer_output = self.sigmoid(layer_input)\n",
    "            self.layer_outputs.append(layer_output)\n",
    "        return self.layer_outputs[-1]\n",
    "\n",
    "    def backward(self, X, y_true):  # 反向传播\n",
    "        output_error = y_true - self.layer_outputs[-1]\n",
    "        output_delta = output_error * self.sigmoid_derivative(self.layer_outputs[-1])\n",
    "\n",
    "        self.weight_gradients = []\n",
    "        self.bias_gradients = []\n",
    "\n",
    "        for i in range(self.num_layers - 2, -1, -1):\n",
    "            # 计算梯度\n",
    "            weight_gradient = np.dot(self.layer_outputs[i].T, output_delta)\n",
    "            bias_gradient = np.sum(output_delta, axis=0, keepdims=True)\n",
    "\n",
    "            # 保存梯度\n",
    "            self.weight_gradients.insert(0, weight_gradient)\n",
    "            self.bias_gradients.insert(0, bias_gradient)\n",
    "\n",
    "            # 更新output_delta用于下一层\n",
    "            if i > 0:\n",
    "                output_delta = output_delta.dot(self.weights[i].T) * self.sigmoid_derivative(self.layer_outputs[i])\n",
    "\n",
    "    def update_weights(self):  # 更新w和b\n",
    "        for i in range(self.num_layers - 1):\n",
    "            self.weights[i] += self.weight_gradients[i] * self.learning_rate\n",
    "            self.biases[i] += np.sum(self.bias_gradients[i], axis=0, keepdims=True) * self.learning_rate\n",
    "\n",
    "    def train(self, X, y_true, epochs=1000):\n",
    "        losses = []\n",
    "        for epoch in range(epochs):\n",
    "            # 前向传播\n",
    "            self.forward(X)\n",
    "\n",
    "            # 计算损失\n",
    "            loss = self.mean_squared_error(y_true, self.layer_outputs[-1])\n",
    "\n",
    "            # 反向传播\n",
    "            self.backward(X, y_true)\n",
    "\n",
    "            # 更新权重和偏置\n",
    "            self.update_weights()\n",
    "            \n",
    "            # 记录损失\n",
    "            losses.append(loss)\n",
    "            \n",
    "        plt.plot(range(epochs), losses)\n",
    "        plt.xlabel('训练次数')\n",
    "        plt.ylabel('损失函数值')\n",
    "        plt.title('训练过程中损失函数变化')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cae41b2a",
   "metadata": {},
   "source": [
    "## 测试代码是否能跑通"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a803914f",
   "metadata": {},
   "source": [
    "#### 1. 回归"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "c5ec22d9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "预测值:\n",
      " [[0.06645521]\n",
      " [0.92875705]\n",
      " [0.92942389]\n",
      " [0.08310243]]\n"
     ]
    }
   ],
   "source": [
    "layer_sizes = [2, 5, 4, 1]  # 输入层2个节点，隐藏层分别有4个和3个节点，输出层1个节点\n",
    "learning_rate = 0.01\n",
    "\n",
    "nn = NeuralNetwork(layer_sizes, learning_rate)\n",
    "\n",
    "# 训练数据集\n",
    "X_train = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])\n",
    "y_train = np.array([[0], [1], [1], [0]])\n",
    "\n",
    "# 训练模型\n",
    "nn.train(X_train, y_train, epochs=100000)\n",
    "\n",
    "# 测试模型\n",
    "test_data = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])\n",
    "predictions = nn.forward(test_data)\n",
    "print(\"\\n预测值:\\n\", predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fba3877b",
   "metadata": {},
   "source": [
    "#### 2. 分类"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "4474a9bf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "预测值:\n",
      " [[9.99653872e-01 3.46128481e-04]\n",
      " [2.98668659e-04 9.99701331e-01]\n",
      " [2.98503990e-04 9.99701496e-01]\n",
      " [9.99681332e-01 3.18668428e-04]]\n"
     ]
    }
   ],
   "source": [
    "layer_sizes = [2, 4, 3, 2]  # 输入层2个节点，隐藏层分别有4个和3个节点，输出层2个节点\n",
    "learning_rate = 0.01\n",
    "\n",
    "nn_clf = NeuralNetwork_clf(layer_sizes, learning_rate)\n",
    "\n",
    "X_train = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])\n",
    "\n",
    "# One-hot编码的标签\n",
    "y_train_one_hot = np.array([[1, 0], [0, 1], [0, 1], [1, 0]])\n",
    "\n",
    "# 训练模型\n",
    "nn_clf.train(X_train, y_train_one_hot, epochs=100000)\n",
    "\n",
    "# 测试模型\n",
    "test_data = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])\n",
    "predictions = nn_clf.forward(test_data)\n",
    "print(\"\\n预测值:\\n\", predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d575fbf",
   "metadata": {},
   "source": [
    "## 尝试实现一个分类任务"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "ff6cff9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_classification\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "# 生成一个二分类数据集\n",
    "X, y = make_classification(n_samples=1000, n_features=10, n_informative=5, random_state=1017)\n",
    "\n",
    "y_onehot = np.zeros((y.shape[0], 2))\n",
    "y_onehot[np.arange(y.shape[0]), y.flatten()] = 1\n",
    "\n",
    "# 将数据集拆分为训练集和测试集\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y_onehot, test_size=0.2, random_state=1017)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "42b77b11",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:94.875%\n",
      "测试集准确率:92.5%\n"
     ]
    }
   ],
   "source": [
    "# 定义层维度\n",
    "layer_dims = [X.shape[1], 5, 3, 2]\n",
    "\n",
    "model = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model.train(X_train, y_train, epochs=1000)\n",
    "\n",
    "train_predictions = model.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "9dfa8e1a",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:96.0%\n",
      "测试集准确率:91.0%\n"
     ]
    }
   ],
   "source": [
    "# 训练神经网络\n",
    "model.train(X_train, y_train, epochs=100000)\n",
    "\n",
    "train_predictions = model.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1529227c",
   "metadata": {},
   "source": [
    "可以发现次数为1000时，模型已经基本训练完成。过多的训练次数可能会使模型过拟合，即出现在训练集上表现变好，而在测试集上表现变差（泛化能力变弱）"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba0650ac",
   "metadata": {},
   "source": [
    "## 探究隐层层数对模型效果的影响"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1b05055",
   "metadata": {},
   "source": [
    "### 隐层层数=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "db1ce314",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:91.875%\n",
      "测试集准确率:92.0%\n"
     ]
    }
   ],
   "source": [
    "# 定义多隐藏层神经网络的层维度\n",
    "layer_dims = [X.shape[1], 5, 2]\n",
    "\n",
    "model_1 = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model_1.train(X_train, y_train, epochs=10000)\n",
    "\n",
    "train_predictions = model_1.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model_1.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1eef31c",
   "metadata": {},
   "source": [
    "### 隐层层数=2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "ecf680cb",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:96.625%\n",
      "测试集准确率:93.0%\n"
     ]
    }
   ],
   "source": [
    "# 定义多隐藏层神经网络的层维度\n",
    "layer_dims = [X.shape[1], 5, 5, 2]\n",
    "\n",
    "model_2 = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model_2.train(X_train, y_train, epochs=10000)\n",
    "\n",
    "train_predictions = model_2.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model_2.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c29da023",
   "metadata": {},
   "source": [
    "### 隐层层数=3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "4be82ac8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:97.25%\n",
      "测试集准确率:95.5%\n"
     ]
    }
   ],
   "source": [
    "# 定义多隐藏层神经网络的层维度\n",
    "layer_dims = [X.shape[1], 5, 5, 5, 2]\n",
    "\n",
    "model_3 = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model_3.train(X_train, y_train, epochs=10000)\n",
    "\n",
    "train_predictions = model_3.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model_3.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5b9428b",
   "metadata": {},
   "source": [
    "### 隐层层数=4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "08067531",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:95.0%\n",
      "测试集准确率:94.5%\n"
     ]
    }
   ],
   "source": [
    "# 定义多隐藏层神经网络的层维度\n",
    "layer_dims = [X.shape[1], 5, 5, 5, 5, 2]\n",
    "\n",
    "model_4 = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model_4.train(X_train, y_train, epochs=10000)\n",
    "\n",
    "train_predictions = model_4.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model_4.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "3f5dbc21",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:48.125%\n",
      "测试集准确率:56.00000000000001%\n"
     ]
    }
   ],
   "source": [
    "# 定义多隐藏层神经网络的层维度\n",
    "layer_dims = [X.shape[1], 5, 5, 5, 5, 5, 2]\n",
    "\n",
    "model_5 = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model_5.train(X_train, y_train, epochs=10000)\n",
    "\n",
    "train_predictions = model_5.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model_5.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15429fdd",
   "metadata": {},
   "source": [
    "* 可以发现刚开始时，随着网络层数的增加，模型的效果得到了很好的提升\n",
    "* 之后，可能是由于过多的层数增加训练难度，使得梯度消失，难已收敛，导致模型效果下降"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39cd1ef4",
   "metadata": {},
   "source": [
    "### 增加隐层中神经元个数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "38bbf6d6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAHECAYAAAA6ZC5mAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAABSTklEQVR4nO3deVxUVeM/8M+sLLIrZiIqilqSZBmJG2nZYupjLj2Z9thmWpb1qzT10XIrJR+zMkvT3BPNLPWr5o5LLmi4gAjuIouoiDDDOgwz5/cHcWNkxxlm9H7er9d9yb3nzp0zB+R+OPfccxVCCAEiIiIimVLauwJERERE9sQwRERERLLGMERERESyxjBEREREssYwRERERLLGMERERESyxjBEREREssYwRERERLLGMEREJHNms9neVSCyK4YhIgd2+vRpmx07KSkJCQkJd3wcIQQKCgoqPaHqdDps3LgRr7/+OrKysu74PQHgyy+/xPbt26u9//Dhw7Ft2zarvHeJ33//HWvXrq32/rm5uSgsLLzj9z137hz27t1b7naTyVSjY8XHx6Nhw4bYs2dPjV5nMpmQmZlpse3GjRvIzc2t0XGIHAHDEJGDMhgM6NWrF4YPH25xgrt161aFJ5xDhw4hIyNDWk9OTkZoaGi5oWft2rXo3r078vLyKqzD5cuXsWXLlkrrmZSUBBcXF2zatMliuxACL7/8Mlq1agVvb28MHToUKSkp2LVrFw4fPgyFQlHhkpOTU+l7AsDs2bOxa9euKvcrsXPnTqSkpEjrP//8MxQKBdRqtcWiVCoxYcKEco9x7tw5i/bdsGEDFi9eXO063HfffXBycqr0s5csY8aMsXjtyZMn8cQTT0Cv1+P333/H+PHjAQAvvPACfvnlFwBA3759MW7cuHLfu6ioCPn5+TCZTCj9FKa2bduiYcOG+OGHHyz2F0KgqKgIeXl55Qbdbdu2oVmzZoiKipK2Pfjgg1iyZEm124PIUTAMETkoJycn7Ny5E3/88Qf+85//SCek5cuX4/HHHwcApKenW/RMvPzyyzh+/Li0vmrVKiQnJ6Np06Ywm83IycmRgtX69evx9ttvw9XVFQBQWFgInU5nUYfJkyfjpZdewvnz5yusp5eXFwDA2dnZYrtCocCLL76IOXPmYMKECQgNDcXOnTsxaNAgODk5AQDS0tIslnnz5gEAtFptle2j1Wqh0Wiq3K+ERqOBWq2W1uvVq4dmzZqhqKjIYhk2bFiZz1Ji+vTp6Nu3r7SuVqtRv379atfh1KlTSElJKfO509LSEBsbi0GDBgEAevfujSFDhli8Njg4GD4+Phg7diy0Wi20Wi1Onz6N3bt348knn0R2djYuXLiAf/3rX+W+97Zt2+Dq6ioFvtLBKyEhAevWrbPYplQqodFoUK9ePcTGxpY53vfff4927dohNDRU2ubi4lLme5Kfn1/t9iGyF3XVuxCRvbRp0wY7d+7EM888g9jYWLRv3x6urq5wd3cHACQmJmLIkCH497//DaD4hF9yMjKZTFiwYAG++OIL1KtXD+fPn0fr1q0tjn/o0CFMmzZNWlepVCgqKpLWv/vuO0RGRuLVV1/FgQMHkJycLJ0oS5Sc7G7duoXExEQolUo0bdoUADBgwAAAwMWLFy3et6SOjRo1stju6elpUQ4Uh7TyLi1pNBoIIcrtRXJ1dYVSqUR+fj6+/fZbODk5QafTYdu2bUhJScGoUaMghJAu8ZVW0WWm7OxsrF+/3iJ8lgSG0oxGI4xGoxQySwsICCizLTExEfPmzcOCBQvQoUMHHD582CJglNi1axd69uwJrVaLqKgoZGRkYP/+/XjppZfw559/Qq1Ww93dHZ07d5ZeYzAYpDr26NEDFy9eRFFREQwGA3x9fTFt2jScPHkSv//+O4Din7dp06bhpZdegsFggBACRqNR+n6WiImJwdatW6vsNTxz5gyeeeYZREREoGvXrpXuS2RPDENEDi4oKAhnz56Fm5sbgOIekZIwUlnPyK+//go3NzcMGzYM8fHx6NevH2JjY9GyZUssWLAAK1euxOHDh6X9zWZzmWDg6emJ7777Dt999x10Oh169+5d4Timkp6MJ554QhrPkp+fDxcXF2mfwsJCmM1mKdz89NNPFscoueRiNBql3qEpU6Zg5syZ5b5neHg4wsPDy2yPi4tDUFAQjEYjDh48CGdnZxQUFODixYvIzs7GG2+8gfz8fOkS3+1KLkGV9vXXXyM3Nxe9e/cuU7ZixQqL9aCgIMTFxZVb5ytXrgAAmjVrBqD40pbZbMaaNWvQp08fizBa2pYtWxATE4P09HTEx8fD29sbGzZskF7boEED6HS6Mj8TK1euxCuvvIJ69eqhRYsWGDduHJYuXYqTJ0/C1dUVWq1WCqUKhQKenp74/fff8eOPP2L79u1o3rx5mbp88MEHAIBevXqVW1egODy++OKLcHZ2RqtWrSrcj8ghCCJyOOfPnxfDhg0TI0aMECNHjhRxcXFS2fLly0VoaKgQQogTJ04IlUollbVs2VLs2bNH5OXlicDAQLF582YhhBDvvvuueOKJJ0RaWprQ6XSiW7duYubMmSI7O1vk5eVVWR+z2SyEECI7O1sUFRVZlOXn5wsAYseOHcJoNErHMxqNAkCZZejQoeLUqVOiS5cuFS6l6zR58mTh5+dXpk4tW7YUkydPtti2Z88eAUCcP3++3P2XLl0qrRuNxkqX0i5duiTc3NzEggULRFpamrQMGDBAvPLKKxbbrly5Iq5cuVJhW44ZM0Y0btxYnD59WgghRJcuXcR3330nhBBi06ZN4rHHHhO5ubnlvjY1NVW0bt1atGnTRnTr1k188skn4urVq8JoNIoGDRqI7777TmRnZ4vs7Gzh5OQktm/fLvLz86XXHzx4UGg0GjFp0iQhhBAff/yxeOKJJ6RyT09PsXTpUhEfHy8aNWokWrVqJVJTUy3qsHDhQul7eTs/Pz8xf/58kZWVJbp06SKaN28ukpKSKmwLIkfBniEiB1QyFsXFxQUzZszAwIEDERQUBABlem/KM3nyZGRmZiI+Ph7R0dFYsWIFjh49imbNmkGr1SI3Nxfx8fH4/PPP8cgjj+DPP/8sc4zExESoVCpoNBoolUo0bNgQbm5u2LRpE5ycnPDMM89Y7F96MHLJZ7h06RK0Wi2GDh0Ks9mMlStXwsnJCY0aNcKBAweq1RYqlapa+5VW+jLe7QoLC6FWq6HVai0GEt/uypUr0uWhkSNHIjg4GCNGjIBCoZD2yc/PR7Nmzcpc7ivPlClT0Lp1a4SHhyMzMxM9evTAqVOnpPZauXIl3nzzTbz99tu4cOECZs2ahfDwcDRp0gQAsHXrVgwfPhwvvfQSmjZtil9//RVCCPTs2RNjxozBzZs3YTKZ4ObmhszMTBgMBrRo0UIa/3T+/Hn06dMHffr0kS6Nms1mi8HRRqMRJpMJDz74IHbv3o1u3brhueeew4kTJ6BSqbBv3z689957GDBggHRp7Xbp6eno0aMHMjMzsWfPHvj7+1fZNkT2xgHURA6oefPmmDNnDr744gsAlgOK8/Pzqxxg3KtXLwwYMAAajQbLly/Hhx9+iAceeAAGgwHvvfcehg4dips3b+Lzzz+XBjPf7r333kPTpk1x//3346GHHpK2f/755xgzZkylQaJEQEAAtFotDh48CK1WKwWH69ev4+LFi0hOTkZKSoq0JCUl4fz58xYDuUuHj+q6PQyZzWaYTCYsWbIEzZs3R1ZWFpydnREREVFmIHPJJb569eoBKL6raunSpdLdZ6VlZmZKly9LKyoqKnPHX2RkJOLj46FSqbBw4UJ8+umnaNiwoVSekpKCpUuXYu7cuXB2dsaqVauk8Vh6vR7z5s3DZ599hi+//BJ5eXno2rUrwsPD8d5772HevHnw8vLCyZMnAQBXr161GLsFFP9MTZ48GStXrpQ+R35+vkW4NhgM0nrbtm2xefNm/Pjjj1Ig/fTTT9G7d298+OGHFbb91KlToVAocOjQoXIvsRE5IvYMEd0FSveOXL9+XbqDqyI9evRAjx49sGnTJri6uuK///0vAODgwYOYO3euNJBYp9PBYDAgJSVF6jEpOYGuW7cOGo0G8+fPx6xZs6Rjz5gxAz179sSmTZsqvHOptO+++w5FRUUQQmD8+PF44403MG/ePCxcuBBOTk4wmUzIzc2Fp6enNG5p0aJFePXVVwFACl2394iJv2/9Lr399oHW165dwyeffILt27fjxo0bCAgIQEREBHx8fKBUKuHt7Y1GjRrhnXfeQU5ODlauXImbN29atPmVK1fKHfhcIioqCtOnTy+3rHRgVKlUUi+QUqnEoEGDLOrevn17PPvsswAgjWMqGf/j4eGBiIgIaDQa7Nu3D1OmTEFMTAzMZjMGDRqE9evX491338Xnn38OoHgKgMDAQIvQrNFo8MEHHyArKws5OTlQq9UIDw+3GCt27do1ODk5SeuhoaEWAXDz5s1QKpVS6Cpt/vz5uHHjBp5//nn88ssv5Y7FInJUDENEd5nTp09X6y/uK1euYNSoUdiyZYsUOrZv3w6DwYB+/fpBqVSioKAAhYWFaN++PfLz89G1a1dpIsPSt5eX7ml56qmn0K1bN2RnZ1dZh7S0NMydOxcBAQFQKBRISkrCyJEjsWfPHsydOxcAsHfvXvTs2bPCyRhNJhNSU1PLPbl+8cUXUu9ZaSWDkBs0aIDjx49j0qRJmDVrFoYNG4bu3buX2T8lJQUNGjSw2FYSZPz8/HD+/Hk4OTlZDE4+dOgQBg4cCKVSiV9++cXibimj0VgmvJVuwx07duCNN97AypUrAQBZWVl4/fXX0a1bN4uem9KaNWtm0WPWtm1b6eu9e/ciJCQEo0ePxqlTp3Do0CF06dKlzDEAoHv37oiJiSm3rLSuXbti7969FkHcw8OjzH6xsbH48MMPcfToUahUKjz//PMMQnTX4WUyorvErVu3ABRPvtehQ4dK901JScGTTz4JLy8vTJs2DQ8++CA6duyIcePGITs7G9evX5cuk4WFheHmzZvIzMys9ozO+/fvx9ChQwFU/CgHs9mMN998E82bN8eoUaMAALNmzUJhYSFu3LiB48ePIz4+HklJSQCKb8M+ffq0xTxJAODm5oauXbtKt8KXLC1btsTkyZMttkVFRaFly5ZSkFGr1YiNjcXo0aPLXA40GAzS1xkZGRaXAkuXazQaBAYGwt/fH40aNUKjRo3QsGFDfP3113jppZfw4Ycf4r///S80Go1U7u/vX+YOKoVCAYPBgJdeeglDhgzBN998gx49egAonqvpwIEDOHLkCJ5//vlyJ8K8dOkSrly5Ajc3Nxw6dAj5+flYtmwZ/Pz80K1bN7i6umLAgAH45ptvsG7dOvTs2bPc78uhQ4dgNBoxePBgPPXUU9It9EIImEwmDBo0CE2aNMHatWurHK914sQJPPLII8jJycGxY8dqNOcSkSNhGCK6C+zfvx+PP/449u3bh+TkZGnwckVBxNfXF/7+/nj00UfRvXt3LFq0CNu2bUO9evVw/PhxtG/fHteuXZP2X7t2LTp16mQREKqi1+tx/Phxad6d2ycq3LJlC3bs2IHFixdLl4eaNGmCgwcPorCwEB06dEBISAhGjRoFk8mE0NBQdOrUqUyPxpgxY8od4F2ejh074sKFCxZBpLzB1EIIPPbYY/Dy8kJmZiZOnjyJkJAQAIC7uzueffbZCscqCSHw7rvv4ty5c/jqq6/w+eefw8vLC926dSt3csLSr3NyckL79u1x4sQJ9O7dGxcuXJB6nQIDA7Fr1y7k5OSU20vm4+ODuXPnQqvV4sSJEzh37hxmzZqF8ePHS5/xo48+wpIlS5CXl4eBAweWW4+SiRcXL16M3Nxc/Pvf/0ZOTg6MRiPeeust7N27F1u3bsX9999f4Wcp8cgjj0gzit8+h1WJ6gz4J7I3hiEiB1YyBmbKlCkYPXo0fvrpJ/Tq1Uu6e8loNFrsX9Ij4uTkhL1792L58uV4+eWXceXKFXz22WdIT0/H4MGD0aFDB9x3333S655++mlcv34dH3/8scXxMjIysHnzZmRlZeHIkSNl6hYaGorXX38dQUFBePTRRy3K+/bti8jIyHJ7sfz8/KRBxps3b4ZKpUJWVhb0en2dzFisUChw+PBhdOzYER9++CGaNm2KJk2aoHv37rh48SK2bdtm0T4lUlNT0bdvX6xatQp//PEH/Pz84OzsjMjISDRr1gyPPvooXn311Uqf+TZhwgT4+/vjjz/+QKtWraQ7wgCgdevWOHr0KBo3blzua8ePH48ffvgBJ0+exKOPPor4+HjExMTgwoULAP4JXB4eHtDr9ZW2gaurK7Zt24asrCyEhISge/fu+PPPP7Fv374yvWS3K/k5E0KgR48eld69t2HDBjz22GNSDyCRI2IYInJg0dHRAICBAweidevWWLVqlcWzpxo2bGgRYEpma46OjsbHH3+Mxx9/HI0aNcLcuXPh6uqKXr16oX379liyZAkUCgXy8/OlgcQ//fQTvv/+e0RGRgIovvspODgYnTp1woYNG/DRRx8hNDQUU6dOxW+//YaEhAR8//33+O2337BixQrExcVh79692LBhg3QiDgsLA2B5C3dGRgZiY2Nx7tw5nDlzxuIy2ZkzZxAbG1vmUll5zGZzte5oy8jIQGJiIm7evGkx5ufMmTPo168fNm7ciIiICDg7OyMoKAjPPvss/vWvf0kBAyh+/tr06dPRtm1bnDlzBvv377cIeW5ubtiyZQu+/fZbbNy4EW3btkVwcLDFY0xMJhP0er1051y7du2wcuVKHDx4EBqNxuKuutK9dqU1aNAAHTt2xPXr19G6dWts3rxZ+n4vWbIE3bt3x6RJk+Dp6YknnngCly5dqrBd9Ho99u7dCw8PDyQlJeHkyZNo3LgxoqOjcenSpUrbtmSW7tt7JrVaLQ4dOoTLly8jMTERly5dQkREBM6ePWtx5xyRw6mT2YyIqFaio6PFY489JvR6vViwYIEYMWJEpfs3bNhQbNq0SaxevVo0a9ZMTJ8+XaSkpAghhJg4caIICwsTeXl54ty5c6J3796ifv36YtSoUdLrX3rpJbFkyRJhNpvFm2++KUaOHCmVmc1m8euvv4r+/fuLli1bCg8PD6FWq8tMqti6desy9QoPDxddu3YVQgixcuVKoVAohIuLi/D09CyzuLm5CS8vryrbpnHjxmLChAlV7vfZZ58JAKJhw4bi0qVL4ujRo6J9+/ZCoVCILl26iPj4eIv9Y2NjRXBwsGjSpIkoKCgQOTk5olWrVsLV1VVMmDBB5OTkVPp+6enp4rPPPhP9+/e32N6pU6dyJ6GsbLl48aL0+uXLl4vnnntOeHh4iAkTJkgTUx47dkx07txZ+Pj4iF9++UUIIcT169dFcHCwqFevnti9e7cQonjCxv/+979i6NChIjg4WKhUKhEYGCgmT54srl+/Lq5evSomTJgg/Pz8BADh4uIiHnroIfHCCy+UmQRy27ZtAkCZtpg4caJQKpUWn0Gr1YpZs2ZV+X0isieFENX404qI7MZsNld6GaK03NxcODs7QwgBpVJp8bqS29BL5sV5//338cADD+CVV14p9y4h8fdzqarz0NSqTJ8+HZs3b8aRI0dQUFAgTeZ4J1xdXTFq1CjMnj270v1KBos/8MAD0oDgadOmoXv37lLP1e0MBgPOnTuHdu3aASi+Vd3Hx6fMHWe2kpKSAn9/f+mxIgCwZ88exMTEYNiwYfDx8ZH2zc3NxVdffYVRo0ZZ1K+goAARERF44403pG0vvvgitFotOnXqhJ49e+KBBx4o895CCMTGxmL//v2Ii4tD165d8Z///KfadS954G2J0o+PIXJUDENEREQka4zrREREJGsMQ0RERCRrDENEREQkawxDREREJGt8NlkVzGYzrl69Cnd391o9PZuIiIjqnhAC2dnZaNy4cZV3NDIMVeHq1avw9/e3dzWIiIioFpKTk9GkSZNK92EYqoK7uzuA4sYsby4WIiIicjx6vR7+/v7SebwyDENVKLk05uHhwTBERER0l6nOEBcOoCYiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGLKjAqPJ3lUgIiKSPYYhOzl9VYcHPt2GCb/H2rsqREREssYwZCff77kAAFh9NNnONSEiIpI3hiEiIiKSNYYhIiIikjWGISIiIpI1hiEiIiKSNYYhIiIikjWGISIiIpI1hiEiIiKSNYYhIiIikjWGISIiIpI1hiEiIiKSNYYhIiIikjWGISIiIpI1hiEiIiKSNYYhIiIikjWGISIiIpI1u4WhjRs3okWLFlCr1ejYsSMSEhIAAKNHj4ZCoZCWwMBA6TVxcXEICQmBt7c3xo4dCyHEHZcRERGRvNklDF28eBGvv/46wsPDkZqaimbNmmH48OEAgGPHjmHLli3IzMxEZmYmTpw4AQAwGAzo27cvOnTogOjoaMTHx2PZsmV3VEZERERklzCUkJCAGTNm4N///jfuu+8+vPPOO4iOjkZRURHi4uIQFhYGLy8veHl5wd3dHQCwdetW6HQ6zJkzBy1btsSMGTOwePHiOyojIiIiUtvjTfv06WOxfvbsWQQGBiI2NhZCCLRv3x6pqal44oknsHDhQjRt2hQxMTEIDQ2Fq6srACA4OBjx8fEAUOuy8hgMBhgMBmldr9db74MTERGRw7H7AOrCwkLMnj0bo0aNQkJCAoKCgrB69WrEx8dDo9Fg5MiRAIpDSUBAgPQ6hUIBlUqFzMzMWpeVZ+bMmfD09JQWf39/m3xuBRQ2OS4RERHVjF16hkqbNGkS3NzcMGLECGg0GgwdOlQqmzdvHlq0aAG9Xg+1Wg0nJyeL1zo7OyMvL6/WZd7e3mXqM2HCBHz00UfSul6vt1kgIiIiIvuzaxjauXMnFixYgKioKGg0mjLlXl5eMJvNSEtLg4+PD+Li4izKs7OzodVqa11WHicnpzLhiYiIiO5ddrtMdunSJQwdOhTz589H27ZtAQAfffQR1q5dK+3z119/QalUwt/fHyEhIYiKipLKEhMTYTAY4OPjU+syIiIiIruEofz8fPTp0wcvvPAC+vXrh5ycHOTk5ODhhx/GxIkTsX//fkRGRmL06NF47bXX4OrqirCwMOh0OqxYsQIAEB4ejp49e0KlUtW6jIiIiMgul8m2b9+OhIQEJCQkYNGiRdL2y5cv48yZM+jXrx/c3d3Rv39/zJgxo7iiajUWLlyIIUOGYOzYsTCZTNi3b98dlREREREpxF02HXNqaiqio6PRuXNn+Pr6WqWsMnq9Hp6entDpdPDw8LDKZwCAd1cdx5ZTaQCAxPDeVjsuERER1ez8bfe7yWrKz88Pfn5+Vi0jIiIi+bL7PENyJXBXdcgRERHdsxiGiIiISNYYhoiIiEjWGIaIiIhI1hiGiIiISNYYhoiIiEjWGIaIiIhI1hiGiIiISNYYhoiIiEjWGIaIiIhI1hiGiIiISNYYhoiIiEjWGIbsRAGFvatAREREYBgiIiIimWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxiyEwFh7yoQERERGIaIiIhI5hiGiIiISNYYhoiIiEjWGIbsRAGFvatAREREYBgiIiIimWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZs1sY2rhxI1q0aAG1Wo2OHTsiISEBABAXF4eQkBB4e3tj7NixEEJIr7FFmd0o7F0BIiIiAuwUhi5evIjXX38d4eHhSE1NRbNmzTB8+HAYDAb07dsXHTp0QHR0NOLj47Fs2TIAsEkZEREREYQdbNq0ScyfP19aj4yMFFqtVqxfv154e3uL3NxcIYQQJ0+eFF26dBFCCJuUVYdOpxMAhE6nu8NPbWnUqmOi2bjNotm4zVY9LhEREdXs/K22RwDr06ePxfrZs2cRGBiImJgYhIaGwtXVFQAQHByM+Ph4ALBJWXkMBgMMBoO0rtfrrfGRiYiIyEHZfQB1YWEhZs+ejVGjRkGv1yMgIEAqUygUUKlUyMzMtElZeWbOnAlPT09p8ff3t8GnJiIiIkdh9zA0adIkuLm5YcSIEVCr1XBycrIod3Z2Rl5enk3KyjNhwgTodDppSU5OtsKnJCIiIkdl1zC0c+dOLFiwABEREdBoNPDx8UF6errFPtnZ2dBqtTYpK4+TkxM8PDwsFptwgBvaiIiIyI5h6NKlSxg6dCjmz5+Ptm3bAgBCQkIQFRUl7ZOYmAiDwQAfHx+blBERERHZJQzl5+ejT58+eOGFF9CvXz/k5OQgJycH3bp1g06nw4oVKwAA4eHh6NmzJ1QqFcLCwqxeRkRERKQQou5nINywYQP69+9fZvvly5dx8uRJDBkyBO7u7jCZTNi3bx+CgoKk11m7rCp6vR6enp7Q6XRWvWT27qrj2HIqDQCQGN7basclIiKimp2/7RKGqpKamoro6Gh07twZvr6+Ni+rDMMQERHR3acm52+7zDNUFT8/P/j5+dVZGREREcmX3W+tJyIiIrInhiEiIiKSNYYhIiIikjWGISIiIpI1hiEiIiKSNYYhIiIikjWGISIiIpI1hiEiIiKSNYYhe1HYuwJEREQEMAwRERGRzDEMERERkawxDBEREZGsMQwRERGRrDEMERERkawxDBEREZGsMQwRERGRrDEMERERkawxDBEREZGsMQwRERGRrDEMERERkawxDBEREZGsMQwRERGRrDEM2YuwdwWIiIgIYBgiIiIimWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGLIXhb0rQERERADDEBEREckcwxARERHJGsMQERERyRrDEBEREckawxARERHJGsMQERERyRrDEBEREckawxARERHJGsMQERERyRrDEBEREckawxARERHJGsMQERERyRrDEBEREckawxARERHJGsMQERERyRrDEBEREclajcLQyZMnq9zHaDTi1VdfrW19iIiIiOqUuiY7v/DCCzh//jyuXr0KIQSUyrJZymw2Y9++fVarIBEREZEt1SgMabVaaDQaDBo0CDk5Obh8+TICAgKQmJiI5s2bS/+WF5KIiIiIHFGtUotCoUBCQgL8/f2RkJCAwMBAi3+FENau5z1HYe8KEBEREYBqhiGDwYAePXrgxo0bmDhxIq5duwagOBSV9y8RERHR3aJaYchkMqFLly5Qq9W4du0acnNzbV2vex77zoiIiBxDtcKQq6srPv/8c/j4+GDx4sVo0aIFzp07B6PRiHPnzqGwsNDi3+r2EGVkZEhjjkqMHj0aCoVCWgIDA6WyuLg4hISEwNvbG2PHjrW4HFfbMiIiIpK3Go0ZMplMAID69evj6aefBgA8++yzMBgMFv9Wx82bN9GnTx+LIAQAx44dw5YtW5CZmYnMzEycOHECQPGlur59+6JDhw6Ijo5GfHw8li1bdkdlRERERBDVlJaWJt59911RVFRU6X4mk0k0adKkyuM99dRT4ptvvhEAxOXLl4UQQhiNRuHu7i6ys7PL7L9+/Xrh7e0tcnNzhRBCnDx5UnTp0uWOyqpDp9MJAEKn01X7NdUxatUx0WzcZtFs3GarHpeIiIhqdv6uVs9QVlYWWrVqBYPBgLNnz+LWrVsVLmlpaRg1alSVx1y4cCE++OADi22xsbEQQqB9+/ZwcXHBc889h6SkJABATEwMQkND4erqCgAIDg5GfHz8HZURERERVSsMeXl5YceOHdBqtXjsscfg6+uLJk2aIDAwEC1btrRYHnzwQUyePLnKY7Zo0aLMtoSEBAQFBWH16tWIj4+HRqPByJEjAQB6vR4BAQHSvgqFAiqVCpmZmbUuK4/BYIBer7dYiIiI6N5V7TFDnTp1wvfff49z585hyJAhcHJywsyZM6WxPSWLXq9HYWFhrSozdOhQREVFISQkBAEBAZg3bx527NgBvV4PtVoNJycni/2dnZ2Rl5dX67LyzJw5E56entLi7+9fq89CREREd4caT7rYpEkTrFy5EsuXL8fWrVuRk5Nji3oBKO6RMpvNSEtLg4+PD9LT0y3Ks7OzodVqa11WngkTJkCn00lLcnKydT8UEREROZRaPzfjX//6FzZs2AA3NzerVeajjz7C2rVrpfW//voLSqUS/v7+CAkJQVRUlFSWmJgIg8EAHx+fWpeVx8nJCR4eHhYLERER3btqHIaKioqwdOnSSvfZvXu3dBt+TbRv3x4TJ07E/v37ERkZidGjR+O1116Dq6srwsLCoNPpsGLFCgBAeHg4evbsCZVKVesyIiIioho9qBUoHoA8ZswYpKenw8/PD23atEH79u2hVhcf6sqVK/j3v/+NNWvWSHMRVdewYcOQkJCAfv36wd3dHf3798eMGTOKK6pWY+HChRgyZAjGjh0Lk8mEffv23VEZERERkUKImk/H3KBBA7z11ltIS0vD5cuXER8fj7CwMLz22mv4+OOPMWDAAISHh9uivkhNTUV0dDQ6d+4MX19fq5RVRq/Xw9PTEzqdzqqXzN6NOI4tsWkAgMTw3lY7LhEREdXs/F3tnqGTJ0+iVatWqFevHnx8fDBz5kypTKfTYcCAAejXrx969eplsyAEAH5+fvDz87NqGREREclXtccMvfvuu6hfvz7CwsKg1+uxZ88eLFy4EMOGDUNQUBAaN26MPXv2IDU1Fb/++qst60xERERkNdXuGTp48CBu3bqFyMhIbNq0Cf3790dubi4++ugjJCQkwN3dHQCwbt06hIWFoXv37jW6HEVERERkD9XuGVqyZAk2bdoEo9EItVqNlJQUDB48GFlZWVi0aBGuXr0KAJg9ezZeeeUVbNmyxWaVJiIiIrKWavcMtW/fHj/88AOKiorg7OyM4cOHw8/PD1999RXc3NzwzTff4OGHHwYAfPvtt2VmfSYiIiJyRNXuGdq7dy+Cg4NRr149nDlzBh4eHmjZsiWEELj//vtx7tw55OXl4ebNm1AoFLas8z2BLUREROQYqh2GTp8+jTNnziA5ORn79+9HYmIizp49iy5duiA/Px9GoxG5ubno1KkTpk+fbss6ExEREVlNtcPQ/Pnz4eLigqeffhodO3ZESkoK3n33XXz55ZfIzMzEww8/DK1Wiy+//BK//vorbt26Zct6ExEREVlFtcPQrFmzUFhYCB8fHzzwwANYtGgRBg0ahODgYPj6+iI6OhrTpk2DRqPBgAEDsGbNGlvWm4iIiMgqqj2A+r///S+USiVSU1PRpk0bPPbYYxgzZgzUajVGjx4NHx8fdO/eHQDw1ltv4f7777dVnYmIiIisplaP45ATWz2Oo0t4JFKz8gHwcRxERETWZpPHcdwuOzsbWq223DvHDAaDNAkjla8kCBEREZF9VXvM0CeffAK9Xg+gOOx4enrC1dUVLi4ucHZ2houLi/S1t7e3zSpMREREZE3VDkOzZ8/Gzz//DACIiYnB9evXkZeXB5PJBLPZjFWrVqGgoABms5lhiIiIiO4a1Q5Dzs7OmD9/PkwmE0aNGoUffvhBmmX6zTffxNixY3Hx4kUA4KSLREREdNeo9pihevXqoXPnztiwYQP++OMPdOzYEb1798b333+PEydO4OjRo7yDjIiIiO461Q5DCoUCo0ePxptvvomGDRvixx9/RF5eHh566CH069cPFy5cgNlshp+fny3rS0RERGRVNbqb7KGHHkLz5s0xdepUaLValNyVv3v3bhgMBvz11184cuSITSpKREREZAs1vrX+l19+gRAC+/fvxxNPPIFNmzbB09MTYWFhePHFF7F7925b1JOIiIjIJqo9gLq0sWPHSrfae3p6YvDgwbh06RKWLFmC9957z9p1JCIiIrKZaoUhs9kMs9kMALh48SIOHz6MPXv2wMPDA2FhYfjss8/Qq1cvqFQqm1aWiIiIyNqqdZmsoKAABQUFAICWLVti//79WLNmDVxcXAAADRs2xODBg7Ft2zbw6R5ERER0N6lWGHJ1dcWNGzekdbPZjBUrVkCr1Uq9QQqFAjExMTCZTNIDW4mIiIgcXbUHULu6ukpfazQabN++3SYVIiIiIqpLtRpATURERHSvuKMwZDabMXz4cIttM2bMwOrVq++oUkRERER1pcZhyGg04tVXXy1+sVKJX375RSoTQmDRokVISEiwXg2JiIiIbKjGYUilUmHjxo3SurOzs/T1mjVrkJWVhQ8++MA6tSMiIiKysRqHIaVSCY1GI62XPKH+1q1bmDBhAmbMmIH69etbr4ZERERENlSrMUMlAajEzZs3MXDgQHTr1g3vvPOOVSpGREREVBeqfWv9mjVroFKp4OTkhMLCQuzatQtubm7IycnBQw89hCFDhmDWrFm2rCsRERGR1VU7DK1btw46nQ4qlQr5+fmYOnUq0tPTUVBQAC8vLzz33HNQq2v83FciIiIiu6pRGCrh6+uLP//8EwDQoEEDzJkzByNGjMAbb7yBzz77zPq1JCIiIrKRGo0ZunbtGgDLMUNKpRKDBw/GX3/9hV9//RXffPONVStIREREZEs1uq7Vq1cv+Pj4lPswVl9fX6xbtw6PP/44nnzySQQHB1utkkRERES2UqOeoUOHDmHYsGHw9vZGr169cO3aNRQVFUnlbdq0wejRozFu3DirV5SIiIjIFmoUhlxcXPDqq6/i9OnTaNeuHXbt2oWcnByLfd5++22cPn26zHYiIiIiR1Sr2780Go10G33nzp0tyry9vbF582a4ubndee2IiIiIbKzaYSgrKwvOzs4wGAxwd3fHlClTYDQaoVAoYDab4eXlhfHjx2Py5Mk4d+4c/u///s+W9SYiIiKyimpfJmvYsCFatWoFPz8/XLhwAfPnz4ezszOcnJzg4uICd3d3nD17FkuWLMHcuXNtWWciIiIiq6l2z1CbNm1w6tQpvPjiixBCQKFQYPLkyRb7DBo0CGPGjEHz5s2tXU8iIiIim6h2GCqZW6j0HEM6nQ6DBg3CI488gkcffRRjx45FUFCQ9WtJREREZCO1elBrCbPZjE6dOsFkMmHRokXo06cPoqKirFU3IiIiIpur0QDqiIgIJCUlSdu8vb0xbdo0aX3Tpk148cUXcfToUbRq1cq6NSUiIiKygWqHoa5du2L37t1o2bIlPDw8IIRAixYtpHJfX18cOXIEI0eOxKRJk/DLL7/YpMJERERE1lTtMBQREQEAuHnzJho0aIB9+/ZBCAG1Wg2TySQ9omPkyJF4+OGHUVBQAGdnZ9vUmoiIiMhKajTpYlFREdq2bYsbN26gadOmeOihh5CQkAAXFxdpn4CAABw+fJhBiIiIiO4K1Q5DQ4YMgbOzM/Ly8vDGG28AAIxGI1544QX4+flJ+7m4uOD999+3fk2JiIiIbKDaYSg0NBROTk7YtGkTOnbsCKB4AHVERAQGDBgg7bdr1y6Eh4dj6dKl1q8tERERkZVVOwyV9PZMnToVI0eOBADk5eXh+++/R6dOnRAcHAwA8PPzw759+2xQVSIiIiLrq/GDWtetWyd97erqii1btqBp06bStj59+qBPnz7WqR0RERGRjdU4DN3+lPqnnnrKapUhIiIiqmt3NAM1ERER0d3OrmEoIyMDAQEBSExMlLbFxcUhJCQE3t7eGDt2rDR/ka3KiIiISN7sFoZu3ryJPn36WAQhg8GAvn37okOHDoiOjkZ8fDyWLVtmszIiIiIiu4WhwYMHY/DgwRbbtm7dCp1Ohzlz5qBly5aYMWMGFi9ebLMyIiIiIruFoYULF+KDDz6w2BYTE4PQ0FC4uroCAIKDgxEfH2+zsvIYDAbo9XqLhYiIiO5ddgtDpR/yWkKv1yMgIEBaVygUUKlUyMzMtElZeWbOnAlPT09p8ff3t8bHJSIiIgflUHeTqdVqODk5WWwreQSILcrKM2HCBOh0OmlJTk62wicjIiIiR+VQYcjHxwfp6ekW27Kzs6HVam1SVh4nJyd4eHhYLERERHTvcqgwFBISgqioKGk9MTERBoMBPj4+NikjIiIicqgwFBYWBp1OhxUrVgAAwsPD0bNnT6hUKpuUERERESmEnWcgVCgUuHz5Mpo3bw4A2LBhA4YMGQJ3d3eYTCbs27cPQUFBNiuril6vh6enJ3Q6nVUvmTUfv0X6OjG8t9WOS0RERDU7f9s9DJUnNTUV0dHR6Ny5M3x9fW1eVhmGISIiortPTc7fNX5Qa13w8/ODn59fnZURERGRfDnUmCEiIiKiusYwRERERLLGMERERESyxjBEREREssYwRERERLLGMERERESyxjBEREREssYwRERERLLGMERERESyxjBEREREssYwRERERLLGMERERESyxjBEREREssYwRERERLLGMERERESyxjBEREREssYwRERERLLGMERERESyxjBEREREssYwZCcvdmhi7yoQERERGIbsRqtm0xMRETkCnpGJiIhI1hiGiIiISNYYhoiIiEjWGIbsRKGwdw2IiIgIYBgiIiIimWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIlljGCIiIiJZYxgiIiIiWWMYIiIiIllzyDA0evRoKBQKaQkMDAQAxMXFISQkBN7e3hg7diyEENJraltmLwoo7F0FIiIigoOGoWPHjmHLli3IzMxEZmYmTpw4AYPBgL59+6JDhw6Ijo5GfHw8li1bBgC1LiMiIiJyuDBUVFSEuLg4hIWFwcvLC15eXnB3d8fWrVuh0+kwZ84ctGzZEjNmzMDixYsBoNZlRERERA4XhmJjYyGEQPv27eHi4oLnnnsOSUlJiImJQWhoKFxdXQEAwcHBiI+PB4Bal5XHYDBAr9dbLERERHTvcrgwlJCQgKCgIKxevRrx8fHQaDQYOXIk9Ho9AgICpP0UCgVUKhUyMzNrXVaemTNnwtPTU1r8/f1t92GJiIjI7hwuDA0dOhRRUVEICQlBQEAA5s2bhx07dsBsNsPJycliX2dnZ+Tl5UGtVteqrDwTJkyATqeTluTkZOt+QCIiInIoDheGbufl5QWz2YxGjRohPT3doiw7OxtarRY+Pj61KiuPk5MTPDw8LBYiIiK6dzlcGProo4+wdu1aaf2vv/6CUqlEu3btEBUVJW1PTEyEwWCAj48PQkJCalVGRERE5HBhqH379pg4cSL279+PyMhIjB49Gq+99hqeeeYZ6HQ6rFixAgAQHh6Onj17QqVSISwsrFZlRERERGp7V+B2w4YNQ0JCAvr16wd3d3f0798fM2bMgFqtxsKFCzFkyBCMHTsWJpMJ+/btA4BalxEREREphCNMx1wDqampiI6ORufOneHr62uVssro9Xp4enpCp9NZdfzQpxvisDLqCgAgMby31Y5LRERENTt/O1zPUFX8/Pzg5+dn1TIiIiKSL4cbM0RERERUlxiG7ETB57QSERE5BIYhO7m7RmoRERHduxiGiIiISNYYhoiIiEjWGIaIiIhI1hiGiIiISNYYhoiIiEjWGIaIiIhI1u66GajvRa/8dATPt7sfapUCJdMPKRUKKKsRVRVQlDtnkbmce/eFAExmIZWbBaDLNyLXUIRbuYXYGncNt3IL8eXAdngppOkdfCIiIqK7B8OQAzhw4SYOXLhp72pIxv12imGIiIhkg5fJqFxxqTp7V4GIiKhOMAzZSZHZsaeg/ivxlr2rQEREVCcYhuwk/qpj97wUFpntXQUiIqI6wTBkJ0qlYz+p1cE7roiIiKyGYchOVA7+2Pry7kYjIiK6FzEM2Ymj9wyZ2DVEREQywTBkJ47eM8QwREREcsEwZCcq9gwRkYMymwX+s/gIPlkXY++qENUJhiE7cfQwpMs32rsKRGQnCdf0+PP8TayNTrF3VYjqBMMQlWtl1BV7V4GI7MTMmTVIZhiGiIisaPmhRCzaf8ne1bgjDj6kkcjq+GwyO+GIHKJ7T4HRhMn/dxoA0P9RPzRwc7JzjYioOtgzZCeC8/gQ3XNKz89VYDTZsSbWY43fVceTMpGebbBCbYhsgz1DDiAxvLf0dV2FJJNZQAAQAigym5FTUIT/i7mKz7ck1Mn7E9HdQYg7u2z2V+ItvLjgMADL33VEjoQ9Q3ZSUeZRKBR1sqhVSmhUSmjVSrhq1Wjo4Yzh3VrgIT8PAMCbXQPqsDWIyFHd6Z9nf56/aZV6ENkSwxBZaOfnBQDwdNHYtyJEd7l75Uo4L+mTHDAM2Ylw0CHUJdMf8dlkRPJV+rIYfxOQHDAM2YmjZo2SX4KOWj8iqlv8XUBywDBEFpR/pyF2jRMR4Li92ETWxDBkJ46aNUrCEB9NRiRfCvxzncxRf1cRWRPDkJ04+l9bjl4/IiIia2EYshNHffYPe4aIqDT2DJEcMAzZiaPeraXkAOpKXc3Kx7Nf78eqI3yQLckDe4lJDhiG7MRRw9A/d5M5Zv3sbebWMzh7PRsT18fZuypENmNxaz1/FZAMMAzZiaNehvrnMpmDVtDO8gvvjedNEVUXfxOQHDAM2Ymj9rwopFvr7VwRorvQvfj/xlF/VxFZE8OQnThqz5BCmoHavvUgIsfAXwUkBwxDduKol6GkAdR1+CvQUGTC1az8Onu/qvzyVxI+XhsDExNhGbp8I9YcTYIuz2jvqjike+UnxmLMkIPe+UpkTQxDdvJ4gI+9q1AufX4RACDxZm6F+ySk6THh91ik6WoXYLILjCgs+uc37PPf/onO4ZGIS9XV6njWNu63U/jteAo2x14tU1b6JCFH768+gfG/n8KoiGP2rgrZkMWki/dMxCOqGMOQnTSvXw8A0DWwgZ1rYmllVPEt43vOple4T69v/8Tqo8l4f/UJAIDRVP0/HXV5RrSbsgNhs/ZI2y6mFwevTeWED3vSFxTZuwpWlXwrD59uiMOVjIqDblX2nSv+uTh4IcNa1SIH56Cd2ERWxTBkJyWDEj1dNXauSfWdv56NbXHXpPW/EjPxwZoTaDVxK2ZvP1vh64QQ+HrnOWyOvYpjSbcAANf0BeXuuyv+OpYcuGzdihMA4D+Lj2Bl1BUM/elImbJ7cZDslYxcvLUiGseTMuvsPe+VduRT60luGIbsJKy1L74d3B6vd25u76pYeOkxfwCAh7O6TNnTX+/H2z9bXh7ZeLK4N2fengsVHjPq0i18u/s83os4UeX7D18RjWmb4xGbklWDWtedu/lcl5iRBwBIybS8vPnHqTQ8Mn0nDl64Waf1KTKZsefMDejybTP+6J2fj2Nn/HUM+OGQTY4vF/dKwCOqDMOQnbTwdUO/9n54rLljjR16pKkXAOuOabqZY6jWfiXjlQDghr56r7mbnL2Wjf8sPoKTyVn2roqFUauOIyvPiP8sLttjZEs/7r+E15f9hZd+PGyT49/J5cDauldiQ+n8c698JqLKMAyRBbWq+EfCaPrnV6DRZMbghVWfsEr+gtTlGWv11+Tqo0k1fk1N2Psv3GFLjuDP8zfxwvcHy5Q5wt1ZdXXz3JFLGThzTY+NJ1MBAGeuZdvkfXgStw52DJEcMAyRBY2qeLBAUaknyT746TZEXbpV5Wu/3nkO+8+l4+FpOzDm11jo8o1YG52M0asrvjx24Ub5J0KTEBi88DAmbThVpsxQZMI7Px+rUXi6kpGLxz7fhe8ruZxXHTW5m0xfYERWXqG0fr2C3q7vdp/Hw9N24PfjKXdUt7tBmi4fLy2MwnPf/Glxx9KdiL+qx5IDl1F020B+e5zE75XgUPoOMt5NRnJQdmAIyZpaWZyPM3ONuKEvgJerFkXV7DKYG/lP0PjteAp+K+fkvu6Y5baec/aXe6yjl28h6lLx8vkL7QAUD+Cetjkevu5O2Bp3DVvjruHlx5sCAPIKi/DFlgT0euh+dG1V9g698K1nkJFbiP9tP4t3ewRW6/PcCbNZIHjKDgDAmenPwVmjqnDfr3aeAwBM+P0UBjzaxOZ1s5WII0mo56RCv/Z+FtuvZuUjPduAh/29kHzrn/FK1pqm4Pm5fwIANGol/hPaTNrOk3jtWYQ6NiPJAMMQWfD+++62+DQ9Hp+x2+rH/+PUtap3Aiz+yl8bnYwugQ0wfEU0rvw9CPh2P/15GauOJGHVkSQkhvfGpxvicPZaNiLe6gi1Slnnk1wWFP3zDLMbegOa1net0/eva1ez8vHf9cW9eP96uLH0WBcA6BweCQDY8WGYxaVKhZUnbYq/ajlP1b3SS2MPHDNEcsPLZGQh2N/L3lUAACw/fEX6+pN1sXju6/1l7oIq7fZb9VdGXcHRxFs4UMd3SJUo3ZtmzXP+nJ3nMOX/TkvrQggUGO3/8NjSd4RVFEJuv1vNSW3tXz+WDW2Xk3jpEHFbBTaeTMWQRVG4lVuIu8kdh0qmUroLMAyRBTcnNXZ9FFYn71XZWKLbZRuKyh0ALYRAalZ+hSdWez32pMhU8/c1FJmx5mgSbpQKduZSocpkFpi7+zyWHUpE0t89ZCNXHsODn23DmqNJuF7B3E11oXQzV9Tmt7eJSmnt6bxvO/4dplAhRLXvhCxPntFy0s4P1pzEoYsZ+GpHxXNyOQprjhkyCcufYSJHxMtkVEZgQ3ckhvcGAPSbdwAxKbZ5TMammDufcfq7yAuYs/MctCr75frUrHzEperwbFAjaVvpWbnLOwEIIcq9TDT+91Nw0ahwv5cz/LxccPxKJqb1ewgJaXqcvqqX9ou6nIGdCdexI/669DoAOP7p01AqAC9XrdU+X01VdL6r7tgza9GoFLiTKYy+3nUec3efR/iAdhj899i0qpQODpfTc/FAI48y+2TZaF4layqdZzecuIp3ures9bGUpX7O/zXvALa83+1OqkZkE7LpGYqLi0NISAi8vb0xduxYu99mfbfY+F5XHJ34lL2rUaE5fw8+LqzgkSAlm5W16CVIySw7Pqn0UUoud3QJj8TIlcekR5kAsHj2mqGobN1CvtiFA+fLv4SXbzThUnou/jx/E7mFJnz8awx+OnAZhy/98wiMT9bFYvrm+DKvfXT6TrSfthO7/g5JtVGb/xulA1/FPUNmiz4GWz/mTXuHl+Hm7j4P4J+gWR2lp6R4Z9VxZBc4fvCpypfbztzR6121//zNXTrQEzkSWYQhg8GAvn37okOHDoiOjkZ8fDyWLVtm72rdNRq6OyMxvDcSw3vjxKdP44v+D9mlHtXtWCgdRHIMxScj93Jm1K7KiaSsMtvUqn9O4TP/SLAo+3RDHGJTspBrKLLoGSpvjqabOYUYuTK6xnWqrrHrYvD1znMWga70VamSGb5vvwx09lo2Aib8UenjVUr8HHUFuYbiS0F5hf9cEqqoB8h42/bSa89/+ycu3Mip8j0rY74tc2qs2FuYX2iC2Sykz1uR25/T1+7vOwpLuxue9WvNy8suGsvvw0OTt2Pi+lO8ZEYORSFk0EWyYcMGvPHGG0hJSYGrqytiYmLw7rvv4sCBA1W+Vq/Xw9PTEzqdDh4eZbu85SzHUIQu4ZE2e5yCtWhVSoueo3paFZw1KiiVCgghIETxidn899elP4+flwuUyuKneCsUKHM32/2ezkjT1X6sTiMP5wqf0+YImni7YOCjTbD/fDpaNHArd7qEygQ19qhRb0BYa190DPBBUGMPKBUK5BtNWH00CXvPpqOeVgWtWgmfelq4aFUwFgmcvW45T1VAg3poc587buUW4mjiP3Njff7CQ3B3Vpd7abKiX4EfrDlZ7nY3JzV+Ht4RrlrV3z87f/8MCSDpVi7e/vl4tT7rf0Kb4X4vZzTycIaLRiXd9aiAZU+mUll87JJtAkL6eSyuP0rtW7zx9jCjVCjgqlXhur4Afl4u8KmnlX7uhRB///v30QVw7EqmRY/Yj//pgIbuTtCqldJ7KxT//L9QoORGgdLrxTNJrY1Oxg97L5bbBl+9+LD0delvjcXXpeJjTTt4q7pjsarDVef9qpovq6pjVOcjVV2PO6tDdepxp21ZVT2aeLuiTSP3ahyl+mpy/pZFGJo6dSqOHDmCP/74A0Dxf/769evj1q2yEwkaDAYYDP/8tazX6+Hv788wVANCCMSn6eGsUSHqUgbiUnVYfTQZDzfxhIeLBvp8I2JSdDU+URIR0b3ptc7NMeVfQVY9Zk3CkCwGUOv1egQEBEjrCoUCKpUKmZmZ8Pb2tth35syZmDp1al1X8Z6iUCgQ1NgTANDS1w0AMHNA8B0ftyS3G4rMUCoUMAuBQpMZWpUSOYYiaWLD/EITisxmeDhrkJlXCKVCAZVSASe1EnmFJhiKzDCazCgyCYten9J/3bpoVcjMLUSRWUAI8fcluuK/ml21ajT2ckZqVj7M5uK7ZYQQKCwyIzUrH54uGvx2PAWdWtSHh4sGKqUC0YmZ2HIqDa92aobtp6/jPg9nXLqZg7BWvtDlG/F8u/uRfCsP8Wl6PNbMGyujruD0VT0+ero1zELARaPC78dTy/SE3K6huxPyC01wc1bfUY9VafbovQpq7AGzABLS7jws93zwPigVxT2ZNVFgNOF4OZdKgeLB2W5OaigVijI9IkqFAv4+LvgrMbPK91ArFXDRqODurIZZ/DMA+59eG6D4KwVKfv4Uin8uGZdc+qzoktM/vUnFUxmU/EzU0xb/XynpvUFJ3ZUKqVfHWGRGtqEILXzrQaMs/j9W0sNauhfp9t6lkq8h/rkUWmgyw9fNCalZltNjhLX2/fuT/aOiv88t5z6q/G/4qv7Er7K8kuNX/doq3MF7V/X+Vb13VX0fVb++ivI7eHFjL+cq3t22ZNEzNG7cOBiNRsyZM0fa5u/vj6ioKPj5Wc6Wy54hIiKiux97hm7j4+ODuLg4i23Z2dnQasvefuzk5AQnJ6e6qhoRERHZmSzuJgsJCUFUVJS0npiYCIPBAB8fHzvWioiIiByBLMJQWFgYdDodVqxYAQAIDw9Hz549oVJV/PBMIiIikgdZXCZTq9VYuHAhhgwZgrFjx8JkMmHfvn32rhYRERE5AFmEIQB44YUXcP78eURHR6Nz587w9fW1d5WIiIjIAcgmDAGAn59fmbvHiIiISN5kMWaIiIiIqCIMQ0RERCRrDENEREQkawxDREREJGsMQ0RERCRrDENEREQkawxDREREJGsMQ0RERCRrspp0sTaEEAAAvV5v55oQERFRdZWct0vO45VhGKpCdnY2AMDf39/ONSEiIqKays7OhqenZ6X7KER1IpOMmc1mXL16Fe7u7lAoFFY9tl6vh7+/P5KTk+Hh4WHVY9M/2M51g+1cN9jOdYPtXHds1dZCCGRnZ6Nx48ZQKisfFcSeoSoolUo0adLEpu/h4eHB/2x1gO1cN9jOdYPtXDfYznXHFm1dVY9QCQ6gJiIiIlljGCIiIiJZYxiyIycnJ0yePBlOTk72rso9je1cN9jOdYPtXDfYznXHEdqaA6iJiIhI1tgzRERERLLGMERERESyxjBEREREssYwZCdxcXEICQmBt7c3xo4dW63pwqnYxo0b0aJFC6jVanTs2BEJCQkAKm/T2pZRseeeew7Lli0DwHa2pfHjx6Nv377SOtvaulauXImmTZvCzc0NPXv2RGJiIgC2szVkZGQgICBAalPANu1qqzZnGLIDg8GAvn37okOHDoiOjkZ8fLx0oqHKXbx4Ea+//jrCw8ORmpqKZs2aYfjw4ZW2aW3LqNiqVauwfft2ALVvS7Zz1eLi4vDDDz/gm2++AcC2traLFy9i4sSJ2LBhA+Lj49GsWTO89tprbGcruHnzJvr06WMRhGzRrjZtc0F1bv369cLb21vk5uYKIYQ4efKk6NKli51rdXfYtGmTmD9/vrQeGRkptFptpW1a2zISIiMjQ9x3332iTZs2YunSpWxnGzGbzaJz587i008/lbaxra3r119/FS+++KK0/ueff4r777+f7WwFTz31lPjmm28EAHH58mUhhG1+fm3Z5uwZsoOYmBiEhobC1dUVABAcHIz4+Hg71+ru0KdPH7z99tvS+tmzZxEYGFhpm9a2jICPP/4Y/fv3R2hoKIDatyXbuXKLFi3CyZMnERAQgM2bN8NoNLKtraxt27aIjIzEiRMnoNPp8P333+Ppp59mO1vBwoUL8cEHH1hss0W72rLNGYbsQK/XIyAgQFpXKBRQqVTIzMy0Y63uPoWFhZg9ezZGjRpVaZvWtkzu9uzZg927d+PLL7+UtrGdrS8nJweTJk1Cq1atkJKSgjlz5iAsLIxtbWVt27bFoEGD8Oijj8LLywtHjhzB7Nmz2c5W0KJFizLbbNGutmxzhiE7UKvVZWbadHZ2Rl5enp1qdHeaNGkS3NzcMGLEiErbtLZlclZQUICRI0di/vz5Fg9OZDtb3++//47c3FxERkbi008/xY4dO5CVlYUlS5awra0oKioKmzZtwpEjR5CdnY2XX34Zzz//PH+mbcQW7WrLNmcYsgMfHx+kp6dbbMvOzoZWq7VTje4+O3fuxIIFCxAREQGNRlNpm9a2TM6mT5+OkJAQ9O7d22I729n6UlJS0LFjR/j4+AAoPokEBwejoKCAbW1Fv/zyCwYPHozHH38cbm5u+Pzzz3Hp0iX+TNuILdrVlm3OMGQHISEhiIqKktYTExNhMBikX4ZUuUuXLmHo0KGYP38+2rZtC6DyNq1tmZxFRERg48aN8PLygpeXFyIiIjBq1CgsX76c7Wxl/v7+yM/Pt9h25coVfPXVV2xrKyoqKsL169el9ezsbOTm5kKtVrOdbcAWv5Nt2uZWGYZNNWI0GoWvr69Yvny5EEKIkSNHij59+ti5VneHvLw88eCDD4q33npLZGdnS0thYWGFbVpZe/N7Ub7k5GRx+fJlaRk4cKD43//+J9LT09nOVpaRkSE8PT3F/PnzRXJysvj222+Fk5OTOH/+PNvailavXi1cXFzEnDlzxKpVq0SPHj1E06ZN+bvDilDqbrLatp292pxhyE7Wr18vXFxcRMOGDUX9+vVFXFycvat0V1i/fr0AUGa5fPlypW1a2zIq9uqrr4qlS5cKIWrflmznih0+fFh07txZuLi4iICAALF+/XohBNvamsxms5gyZYpo2rSp0Gg04pFHHhHR0dFCCLaztZQOQ0LYpl1t1eZ8ar0dpaamIjo6Gp07d4avr6+9q3NPqKxNa1tGZbGd6w7bum6wnW3DFu1qizZnGCIiIiJZ4wBqIiIikjWGISIiIpI1hiEiIiKSNYYhIiIikjWGISKShZreK3Lp0iUYjUYb1YaIHAnDEBE5pMOHD6NXr17IycmpdL+ioiLodDpp/dChQ1i7dm2Z/fr27YudO3eWe4yzZ8+W2fbOO+/gs88+q/B9TSYTpk+fDoPBgDfeeAOzZ8/GsWPH8NNPPwEAOnfujOPHj1dadyJyDAxDROSQQkND0aJFCzz99NPIzs4GAAwcOBC///47hBDYsGEDAOD48ePo16+f9Lovv/wSx44dAwAYDAYIIXDz5k0cPHgQHTt2BFAcoAoLCwEAhYWFeOaZZ7BgwQLpGBkZGYiKisL/+3//r8L6qVQqXL9+HRMmTIBWq4VGo8EPP/wAlUqFjIwMREdHIzAw0JpNQkQ2wjBERA5JoVBg3rx5CAoKksKNWq2GEAIKhQKDBg0CAGi1WqjVagBAUlIS4uLiMHnyZFy7dg1du3ZFs2bNpFASHByM5s2bo0WLFnj11Vel10dERGDMmDE4evQoAGDZsmXIzc1FUFAQGjRoIC2+vr5ITEwEUPyA1dDQUDRv3hypqak4e/YsjEYjtFotIiMj8fjjj8PDwwNAcfjiJTcix8UwREQOp7CwEDk5OTAYDFiwYAG6d+8OAHBycpL2KQlApc2ePRvh4eHQaDTo2LEjZs2ahaSkJLRt2xZbtmxBYmIiEhMTkZSUhNWrV0uv69KlCz755BOMHz8e+fn5+Oqrr3Dw4EHcvHlTWhYsWABfX180b94cAJCWloZdu3YhIiICmzdvRlxcHJydnbFr1y5s2rQJ586dQ2BgIOrXrw9fX1+sWLHCpm1GRLXHMEREDmfz5s1o2bIlGjZsiDFjxkjbTSZThQOh9+7di/Xr1+PWrVt466230LZtW8ydOxeBgYGIjo7GiBEjEBAQALPZXO7rx48fj/Xr12PSpEno2rUrOnbsiPDwcHzxxRcAgI0bN1pcjgsJCcHw4cORlZWFp556Cq1bt0ZWVhbmzp2LjRs3IiIiAhcuXMBbb72FcePG4c0337RiCxGRNZX904qIyM4GDBiAAQMGYMqUKcjNzZW2FxQUQKVSlfsaFxcXPPHEE7hy5QrWrFmD2NhYtG7dGuPGjYOfnx/ef/99ODs7Q6ks/29ArVYLlUqFp59+Gm3btgVQ/Aykkt6oAQMGIDg4WNp/+fLl+N///ocNGzbghx9+QOvWrXHx4kW88847yMnJwYULF9CzZ0+kpaUhKCjIWk1DRDbAniEicmilw09GRgbc3d3L3a9jx474+eefkZ+fj08++QStW7dGSkoKVq1aJY0PKlFYWCgNoC5x5swZPPLII3jyySfRtGlTAMVhqOROtf79+6Nly5bS/gMHDsSBAweQnp6Ov/76C++88w6+/vpruLq64qOPPsKBAwcAAJcvX0abNm3uvCGIyGbYM0REd43Tp0+jRYsWFZbv27cP+/btQ1RUFAoKCrB7927k5eWhQ4cOAIpDUGBgIAwGAz755BOMHj1aeu3UqVPRtWtXaLVaAMWDng8cOAAXFxcUFRWVGaO0YMECfPfdd0hPT4eXlxfatGmDtLQ0TJkyBSNHjsQDDzyArKwsxMfHo3379tZvDCKyGj61nogc1pQpU1BQUIAuXbqgdevW6N69O65evQqFQgFnZ2cUFBTg5MmTGDNmDH777TcEBASgSZMmMBqNEEJg//798Pb2xrVr1+Dv7w9nZ2fk5ubiypUrFqHq6NGj6Nu3L86cOQNvb28AwI8//ohNmzbB19cXgYGBmDhxYpn6ZWRkoEOHDoiJiYFWq0WrVq1w+PBh+Pv7Y+jQoUhLS4OnpyfWr19fZ21GRDXHniEiclhCCGzcuBG7du1Cz5490b9/fygUCgDFPTeleXp6Yv78+WjatCnat28Po9EIjUaDSZMm4cSJE9ixYwcA4ODBgxgyZAhOnDgBX19f5Obm4rXXXsPUqVOlIHTo0CF89tln2LVrF3x9ffHoo4/C3d0d77//vsV7xsbGokmTJnjmmWdQr149PPzww/Dz8wMAvPXWW+jRo4c0HxIROS6OGSIih3X58mWkp6fjf//7H3788UfpspbJZELXrl0BAEajESaTCUajEc2aNcO+ffvw/PPPo02bNpg9ezbWrVuH5cuXAygOV2FhYejduzdee+01AMD8+fOh1Wrx1ltvwWw247vvvkOfPn2wePFitGvXDo0aNcL27dsxdepUDB06FFevXpXq16NHD0RGRqJz587SQO+3334b58+fxwcffIDnnnsOH3/8sTQ3ERE5JoYhInJYSqUSK1asQFJSEkaMGIEHH3wQQPGg6r179wIonmW6sLAQkZGRePbZZ5GcnIwvvvgCW7duxaJFi7Bjxw5cuHAB77//Pnx8fAAUz1J9/vx5nD9/Hh9//DE2b94MlUqF559/HvPmzcPOnTvRp08fqR7t2rVDVFQUzp07h1deeQVCCFy6dAkTJkxAu3bt4Orqin379mHdunXw9PREt27dMGHCBGzduhUDBw7EI488UuGjQIjI/jhmiIjuCSWP2HB1dZW2Xb9+Hffddx8SEhLw888/Y9CgQXjkkUcAAGazucxt9qmpqWjQoIHF5I6lmUwm6HQ6+Pj4ID8/H4sXL8bgwYPRoEEDaZ8lS5agd+/euO+++6Rt27dvx5NPPgmNRmPNj0xEVsIwRERERLLGy2REREQkawxDREREJGsMQ0RERCRrDENEREQkawxDREREJGsMQ0RERCRrDENEREQkawxDREREJGsMQ0RERCRr/x/sxnKV+rNOdAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:99.5%\n",
      "测试集准确率:97.0%\n"
     ]
    }
   ],
   "source": [
    "# 定义多隐藏层神经网络的层维度\n",
    "layer_dims = [X.shape[1], 10, 10, 10, 2]\n",
    "\n",
    "model_6 = NeuralNetwork_clf(layer_dims)\n",
    "\n",
    "model_6.train(X_train, y_train, epochs=10000)\n",
    "\n",
    "train_predictions = model_6.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "test_predictions = model_6.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8bd2013b",
   "metadata": {},
   "source": [
    "模型表现得到了提高"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48cf451b",
   "metadata": {},
   "source": [
    "## 使用 sklearn 多层感知机算法求解手写数字识别的问题"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b446cbd",
   "metadata": {},
   "source": [
    "* 多层感知机（Multilayer Perceptron, MLP）是一种基于神经网络的模型，它包含多个隐藏层，每个隐藏层包含多个神经元，每个神经元与前一层的所有神经元相连。在训练过程中，多层感知机通过反向传播算法来更新权重和偏置，以最小化目标函数。\n",
    "\n",
    "* 神经网络是一种模拟生物神经系统的计算模型，它由大量的神经元和连接它们的权重组成。神经元接收输入信号，对其进行处理，并将输出信号传递给下一个神经元。神经网络也可以包含多个隐藏层，但不限于此，可以是一个或多个隐藏层。在训练过程中，神经网络使用反向传播算法来更新权重和偏置，以最小化目标函数。\n",
    "\n",
    "* 因此，多层感知机是一种特定类型的神经网络，它包含多个隐藏层。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "a1b5ffd4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "准确度: 0.95\n"
     ]
    }
   ],
   "source": [
    "# 导入必要的库\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn import datasets\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "# 加载手写数字数据集\n",
    "digits = datasets.load_digits()\n",
    "X = digits.data\n",
    "y = digits.target\n",
    "\n",
    "# 将数据集拆分为训练集和测试集\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=1017)\n",
    "\n",
    "# 创建多层感知机模型\n",
    "mlp = MLPClassifier(hidden_layer_sizes=(100,), max_iter=500, activation='relu', solver='adam', random_state=1017)\n",
    "\n",
    "mlp.fit(X_train, y_train)\n",
    "\n",
    "y_pred = mlp.predict(X_test)\n",
    "\n",
    "# 计算准确度\n",
    "accuracy = accuracy_score(y_test, y_pred)\n",
    "print(f'准确度: {accuracy:.2f}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "7ebff607",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "\n",
    "# 计算混淆矩阵\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "\n",
    "# 使用Seaborn库绘制热力图\n",
    "plt.figure(figsize=(8, 6))\n",
    "sns.heatmap(cm, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=digits.target_names, yticklabels=digits.target_names)\n",
    "plt.title('混淆矩阵')\n",
    "plt.xlabel('预测值')\n",
    "plt.ylabel('真实值')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6250f649",
   "metadata": {},
   "source": [
    "### 尝试使用自己编写的神经网络算法进行训练"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "2321f4ce",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "训练集准确率:82.08768267223383%\n",
      "测试集准确率:81.55555555555556%\n"
     ]
    }
   ],
   "source": [
    "# 定义层维度\n",
    "layer_dims = [X.shape[1], 3, 3, 10]\n",
    "\n",
    "model_shouxie = NeuralNetwork_clf(layer_dims, 0.1)\n",
    "\n",
    "model_shouxie.train(X_train, y_train, epochs=100000)\n",
    "\n",
    "# 在训练集上进行预测\n",
    "train_predictions = model_shouxie.forward(X_train)\n",
    "train_predictions = (train_predictions > 0.5).astype(int)\n",
    "\n",
    "# 在测试集上进行预测\n",
    "test_predictions = model_shouxie.forward(X_test)\n",
    "test_predictions = (test_predictions > 0.5).astype(int)\n",
    "\n",
    "# 计算准确率\n",
    "train_accuracy = accuracy_score(y_train.flatten(), train_predictions.flatten())\n",
    "test_accuracy = accuracy_score(y_test.flatten(), test_predictions.flatten())\n",
    "\n",
    "print(f\"训练集准确率:{train_accuracy*100}%\")\n",
    "print(f\"测试集准确率:{test_accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39f8db83",
   "metadata": {},
   "source": [
    "模型训练效果明显不好，为什么准确率还这么高？"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "548034c6",
   "metadata": {},
   "source": [
    "回看代码，发现上述计算准确率的方法只适用于2分类问题的准确率计算，更改计算方式后："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "8c95e94a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "测试集准确率:7.777777777777778%\n"
     ]
    }
   ],
   "source": [
    "y_pred_classes = np.argmax(test_predictions, axis=1)\n",
    "y_true_classes = np.argmax(y_test, axis=1)\n",
    "\n",
    "# 计算准确分类的样本数\n",
    "correct_predictions = np.sum(y_pred_classes == y_true_classes)\n",
    "\n",
    "# 计算准确率\n",
    "accuracy = correct_predictions / len(y_test)\n",
    "print(f\"测试集准确率:{accuracy*100}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8da72da",
   "metadata": {},
   "source": [
    "自己编写的神经网络算法计算速度太慢，故后续没有继续尝试。。。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84e6b8f3",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}