使用Python实现卷积神经网络（CNN）来识别手写数字，可借助深度学习框架如TensorFlow和Keras。以下是详细步骤及示例代码： ### 1. 导入必要的库 ```python import tensorflow as tf from tensorflow import keras import numpy as np import matplotlib.pyplot as plt ``` ### 2. 加载和预处理数据集 使用MNIST数据集，该数据集包含60,000个训练样本和10,000个测试样本，每个样本是一个28x28的灰度图像，代表一个手写数字。 ```python # 加载MNIST数据集 mnist = keras.datasets.mnist (train_images, train_labels), (test_images, test_labels) = mnist.load_data() # 数据预处理 train_images = train_images.reshape((60000, 28, 28, 1)) train_images = train_images / 255.0 test_images = test_images.reshape((10000, 28, 28, 1)) test_images = test_images / 255.0 ``` ### 3. 构建卷积神经网络模型 ```python model = keras.Sequential([ # 卷积层 keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), # 池化层 keras.layers.MaxPooling2D((2, 2)), # 卷积层 keras.layers.Conv2D(64, (3, 3), activation='relu'), # 池化层 keras.layers.MaxPooling2D((2, 2)), # 展平层 keras.layers.Flatten(), # 全连接层 keras.layers.Dense(64, activation='relu'), # 输出层 keras.layers.Dense(10, activation='softmax') ]) ``` ### 4. 编译模型 ```python model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy']) ``` ### 5. 训练模型 ```python model.fit(train_images, train_labels, epochs=5) ``` ### 6. 评估模型 ```python test_loss, test_acc = model.evaluate(test_images, test_labels) print(f"Test accuracy: {test_acc}") ``` ### 7. 进行预测 ```python predictions = model.predict(test_images) ``` ### 完整代码示例 ```python import tensorflow as tf from tensorflow import keras import numpy as np import matplotlib.pyplot as plt # 加载MNIST数据集 mnist = keras.datasets.mnist (train_images, train_labels), (test_images, test_labels) = mnist.load_data() # 数据预处理 train_images = train_images.reshape((60000, 28, 28, 1)) train_images = train_images / 255.0 test_images = test_images.reshape((10000, 28, 28, 1)) test_images = test_images / 255.0 # 构建卷积神经网络模型 model = keras.Sequential([ keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), keras.layers.MaxPooling2D((2, 2)), keras.layers.Conv2D(64, (3, 3), activation='relu'), keras.layers.MaxPooling2D((2, 2)), keras.layers.Flatten(), keras.layers.Dense(64, activation='relu'), keras.layers.Dense(10, activation='softmax') ]) # 编译模型 model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy']) # 训练模型 model.fit(train_images, train_labels, epochs=5) # 评估模型 test_loss, test_acc = model.evaluate(test_images, test_labels) print(f"Test accuracy: {test_acc}") # 进行预测 predictions = model.predict(test_images) ```