tensorflow2逻辑回归（三）

数据集

MNIST，它包含了 0~9 共 10 种数字的手写图片，每种数字一共有 7000 张图片， 采集自不同书写风格的真实手写图片， 一共 70000 张图片。

考虑到手写数字图片包含的信息比较简单， 每张图片均被缩放到28 × 28的大小，同时
只保留了灰度信息。

单层网络实现代码

import numpy as np
import tensorflow as tf

# Parameters
learning_rate = 0.001
training_epochs = 6
batch_size = 600

# Import MNIST data
(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()

# tf.reshape(x_train, [-1, 784] 打平操作， [b, 28, 28] => [b, 784]
train_dataset = (
    tf.data.Dataset.from_tensor_slices((tf.reshape(x_train, [-1, 784]), y_train))
    .batch(batch_size)
    .shuffle(1000)
)

# x缩放到0-1；y独热编码
train_dataset = (
    train_dataset.map(lambda x, y:
                      (tf.divide(tf.cast(x, tf.float32), 255.0),
                       tf.reshape(tf.one_hot(y, 10), (-1, 10))))
)

# Set model weights
W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))

# Construct model
model = lambda x: tf.nn.softmax(tf.matmul(x, W) + b) # Softmax
# Minimize error using cross entropy
compute_loss = lambda true, pred: tf.reduce_mean(tf.reduce_sum(tf.losses.binary_crossentropy(true, pred), axis=-1))
# caculate accuracy
compute_accuracy = lambda true, pred: tf.reduce_mean(tf.keras.metrics.categorical_accuracy(true, pred))
# Gradient Descent
optimizer = tf.optimizers.Adam(learning_rate)

for epoch in range(training_epochs):
    for i, (x_, y_) in enumerate(train_dataset):
        with tf.GradientTape() as tape:
            pred = model(x_)
            loss = compute_loss(y_, pred)
        acc = compute_accuracy(y_, pred)
        grads = tape.gradient(loss, [W, b])
        optimizer.apply_gradients(zip(grads, [W, b]))
    print("=> loss %.2f acc %.2f" %(loss.numpy(), acc.numpy()))

高级API实现三层网络结构

import numpy as np
import tensorflow as tf

# Import MNIST data
(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()

model2 = tf.keras.models.Sequential([
    tf.keras.layers.Flatten(input_shape=(28, 28)),
    tf.keras.layers.Dense(256, activation='relu'),
    tf.keras.layers.Dense(128, activation='relu'),
    #tf.keras.layers.Dropout(0.2),
    tf.keras.layers.Dense(10, activation="softmax")
])

model2.compile(optimizer="adam",
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])

model2.fit(x_train, y_train, epochs=10)