本文共 4319 字，大约阅读时间需要 14 分钟。
from __future__ import print_functionimport numpy as npfrom keras.datasets import mnistfrom keras.models import Sequentialfrom keras.layers.core import Dense, Dropout, Activation, Flattenfrom keras.optimizers import RMSpropfrom keras.utils import np_utilsfrom keras.layers.convolutional import Conv2D, MaxPooling2Dimport matplotlib.pyplot as pltnp.random.seed(1671) # 为 reproducibility 设置随机种子# 网络和训练配置NB_EPOCH = 20BATCH_SIZE = 128VERBOSE = 1NB_CLASSES = 10 # 输出数量等于数字数量OPTIMIZER = RMSprop()N_HIDDEN = 128VALIDATION_SPLIT = 0.2DROPOUT = 0.3# 数据预处理：分割训练集和测试集(X_train, y_train), (X_test, y_test) = mnist.load_data()# 数据形状调整RESHAPED = 784X_train = X_train.reshape(60000, 28, 28, 1)X_test = X_test.reshape(10000, 28, 28, 1)X_train = X_train.astype('float32')X_test = X_test.astype('float32')# 数据归一化X_train /= 255X_test /= 255print(X_train.shape[0], '训练样本数量')print(X_test.shape[0], '测试样本数量')# 类别转换为二元分类矩阵Y_train = np_utils.to_categorical(y_train, NB_CLASSES)Y_test = np_utils.to_categorical(y_test, NB_CLASSES)# 定义模型model = Sequential()model.add(Conv2D(32, kernel_size=5, padding='same', input_shape=(28, 28, 1)))model.add(Activation('relu'))model.add(MaxPooling2D(pool_size=(2, 2)))model.add(Dropout(0.25))model.add(Conv2D(64, kernel_size=5, padding='same'))model.add(Activation('relu'))model.add(MaxPooling2D(pool_size=(2, 2)))model.add(Dropout(0.25))model.add(Flatten())model.add(Dense(1024))model.add(Activation('relu'))model.add(Dropout(0.5))model.add(Dense(NB_CLASSES))model.add(Activation('softmax'))model.summary()model.compile(loss='categorical_crossentropy', optimizer=OPTIMIZER, metrics=['accuracy'])# 训练模型history = model.fit(X_train, Y_train, batch_size=BATCH_SIZE, epochs=6, verbose=VERBOSE, validation_split=VALIDATION_SPLIT)# 评估模型性能score = model.evaluate(X_test, Y_test, verbose=VERBOSE)print("\n测试准确率:", score[1])# 可视化训练历史plt.plot(history.history['acc'])plt.plot(history.history['val_acc'])plt.title('模型准确率曲线')plt.ylabel('准确率')plt.xlabel('训练轮次')plt.legend(['训练', '验证'], loc='upper left')plt.show()plt.plot(history.history['loss'])plt.plot(history.history['val_loss'])plt.title('模型损失曲线')plt.ylabel('损失值')plt.xlabel('训练轮次')plt.legend(['训练', '验证'], loc='upper left')plt.show()模型总结
_______________________________________________________________________
Layer (type) Output Shape Param #
===================================================================
conv2d_1 (Conv2D) (None, 28, 28, 32) 832
_______________________________________________________________
activation_1 (Activation) (None, 28, 28, 32) 0
_______________________________________________________________
max_pooling2d_1 (MaxPooling2D) (None, 14, 14, 32) 0
_______________________________________________________________
dropout_1 (Dropout) (None, 14, 14, 32) 0
_______________________________________________________________
conv2d_2 (Conv2D) (None, 14, 14, 64) 512
_______________________________________________________________
activation_2 (Activation) (None, 14, 14, 64) 0
_______________________________________________________________
max_pooling2d_2 (MaxPooling2D) (None, 7, 7, 64) 0
_______________________________________________________________
dropout_2 (Dropout) (None, 7, 7, 64) 0
_______________________________________________________________
flatten_1 (Flatten) (None, 3136) 0
_______________________________________________________________
dense_1 (Dense) (None, 1024) 3212288
_______________________________________________________________
activation_3 (Activation) (None, 1024) 0
_______________________________________________________________
dropout_3 (Dropout) (None, 1024) 0
_______________________________________________________________
dense_2 (Dense) (None, 10) 10250
_______________________________________________________________
activation_4 (Activation) (None, 10) 0
===================================================================
总参数数量：3,274,634
转载地址：http://afpg.baihongyu.com/