keras对数据集cifar100实现卷积神经网络识别

 

    model = keras.models.Sequential([
        # 卷积层1； 32个5*5*3的filter
        keras.layers.Conv2D(32, kernel_size=5, strides=1, padding="same", data_format="channels_last", activation=tf.compat.v1.nn.relu),
        # 池化层1;2*2窗口
        keras.layers.MaxPool2D(pool_size=2, strides=2, padding="same"),
        # 卷积层2：64个5*5*32的filter
        keras.layers.Conv2D(64, kernel_size=5, strides=1, padding="same", data_format="channels_last",
                            activation=tf.compat.v1.nn.relu),
        # 池化层2：2*2窗口， [None, 8, 8, 64]
        keras.layers.MaxPool2D(pool_size=2, strides=2, padding="same"),
        # [None, 8, 8, 64]--->[None, 8*8*64]
        keras.layers.Flatten(),
        # 全连接层
        # 1024个神经元网络层
        keras.layers.Dense(1024, activation=tf.compat.v1.nn.relu),
        # 100个神经元网络
        keras.layers.Dense(100, activation=tf.compat.v1.nn.softmax)
    ])

完整代码：

from tensorflow.keras.datasets import cifar100
from tensorflow import keras
import tensorflow as tf
tf.compat.v1.disable_eager_execution()
class CNNMnist(object):
    # 定义模型
    model = keras.models.Sequential([
        # 卷积层1； 32个5*5*3的filter
        keras.layers.Conv2D(32, kernel_size=5, strides=1, padding="same", data_format="channels_last", activation=tf.compat.v1.nn.relu),
        # 池化层1;2*2窗口
        keras.layers.MaxPool2D(pool_size=2, strides=2, padding="same"),
        # 卷积层2：64个5*5*32的filter
        keras.layers.Conv2D(64, kernel_size=5, strides=1, padding="same", data_format="channels_last",
                            activation=tf.compat.v1.nn.relu),
        # 池化层2：2*2窗口， [None, 8, 8, 64]
        keras.layers.MaxPool2D(pool_size=2, strides=2, padding="same"),
        # [None, 8, 8, 64]--->[None, 8*8*64]
        keras.layers.Flatten(),
        # 全连接层
        # 1024个神经元网络层
        keras.layers.Dense(1024, activation=tf.compat.v1.nn.relu),
        # 100个神经元网络
        keras.layers.Dense(100, activation=tf.compat.v1.nn.softmax)
    ])
    def __init__(self):
        # 获取数据
        (self.x_train, self.y_train), (self.x_test, self.y_test) = cifar100.load_data()
        # 进行数据归一化
        self.x_train = self.x_train.astype('float32') / 255.0
        self.x_test = self.x_test.astype('float32') / 255.0
    # 编译
    def compile(self):
        CNNMnist.model.compile(optimizer=keras.optimizers.Adam(),
                               loss=keras.losses.sparse_categorical_crossentropy,
                               metrics=['accuracy'])
        return None
    # 训练
    def fit(self):
        CNNMnist.model.fit(self.x_train, self.y_train, epochs=1, batch_size=32)
        return None
    def evaluate(self):
        test_loss, test_acc = CNNMnist.model.evaluate(self.x_test, self.y_test)
        print(test_loss, test_acc)
        return None
if __name__ == "__main__":
    cnn = CNNMnist()
    cnn.compile()
    cnn.fit()
    print(cnn.model.summary())

运行结果如下：