import tensorflow as tffrom tensorflow import kerasimport matplotlib.pyplot as plt%matplotlib inlineimport numpy as np
# 下载数据集并划分为目标集和测试集(train_image,train_lable),(test_image,test_label) = tf.keras.datasets.fashion_mnist.load_data()
train_image.shape
# 在最后一个维度扩张，扩张成一个4维数据train_image = np.expand_dims(train_image,-1)
train_image.shape # (None,hight,witch,chanal)1表示黑白3彩色
test_image = np.expand_dims(test_image,-1)
# 建立模型model = tf.keras.Sequential()# 顺序模型model.add(tf.keras.layers.Conv2D(32,(3,3),input_shape=train_image.shape[1:],activation="relu"))# 建立卷积层#每层建立32个卷积核，卷积核大小（3*3） #输入图片的形状如（60000，28，28，1除去第0位的图片个数）就是28*28*1# 激活函数relumodel.add(tf.keras.layers.MaxPool2D())# 最大池化默认2*2形状model.add(tf.keras.layers.Conv2D(64,(3,3),activation="relu"))# 再次添加卷积层2的n次方形式添加卷积核model.add(tf.keras.layers.GlobalAveragePooling2D())# 全局平均值池化model.add(tf.keras.layers.Dense(10,activation="softmax"))# 输出
model.summary()
# 训练模型pile(optimizer="adam",loss="sparse_categorical_crossentropy",metrics=["acc"])
history = model.fit(train_image,train_lable,epochs=30,validation_data=(test_image,test_label))
history.history.keys()
# 正确率（通过绘图得train数据得分不高未达到拟合 test数据过拟合）plt.plot(history.epoch,history.history.get("acc"),label="acc")plt.plot(history.epoch,history.history.get("val_acc"),label="val_acc")
# 误差plt.plot(history.epoch,history.history.get("loss"),label="loss")plt.plot(history.epoch,history.history.get("val_loss"),label="val_loss")
优化模型
# CNN优化增加卷积层及抑制拟合：增大测试训练集隐藏单元数增大拟合，降低抑制数据拟合# 建立模型model = tf.keras.Sequential()model.add(tf.keras.layers.Conv2D(64,(3,3),input_shape=train_image.shape[1:],activation="relu",padding="same"))model.add(tf.keras.layers.Conv2D(64,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.Conv2D(64,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.MaxPool2D())model.add(tf.keras.layers.Dropout(0.5))model.add(tf.keras.layers.Conv2D(128,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.Conv2D(128,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.MaxPool2D())model.add(tf.keras.layers.Dropout(0.5))model.add(tf.keras.layers.Conv2D(256,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.Conv2D(256,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.MaxPool2D())model.add(tf.keras.layers.Dropout(0.5))model.add(tf.keras.layers.Conv2D(512,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.Conv2D(512,(3,3),activation="relu",padding="same"))model.add(tf.keras.layers.Dropout(0.5))model.add(tf.keras.layers.GlobalAveragePooling2D())model.add(tf.keras.layers.Dense(256,activation="relu"))model.add(tf.keras.layers.Dense(10,activation="softmax"))
