多层感知机（multi-layer perceptron）实现手写体分类（TensorFlow）

#multi_layer  Perceptron by ffzhang
import numpy as np
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
import time
import os
os.environ["CUDA_VISIBLE_DEVICES"]='2'

mnist = input_data.read_data_sets('data/mnist',one_hot=True)

mnist.train.images.shape
mnist.train.labels.shape

#define placeholder to save the training data
X = tf.placeholder(tf.float32,[None,784],name='X_placeholder')
Y = tf.placeholder(tf.float32,[None,10],name='Y_placeholder')

#define parameters
n_hidden_1 =256
n_hidden_2 =256
n_input =784
n_classes =10

weights ={
    'h1': tf.Variable(tf.random_normal([n_input,n_hidden_1]),name='W1'),
    'h2': tf.Variable(tf.random_normal([n_hidden_1,n_hidden_2]),name='W2'),
    'out': tf.Variable(tf.random_normal([n_hidden_2,n_classes]),name='W')
}

biases ={
    'b1': tf.Variable(tf.random_normal([n_hidden_1]),name='b1'),
    'b2': tf.Variable(tf.random_normal([n_hidden_2]),name='b2'),
    'out': tf.Variable(tf.random_normal([n_classes]),name='bias')
}

#define graph for this network
def multilayer_perceptron(x,weights,biases):
     layer_1 = tf.add(tf.matmul(x,weights['h1']),biases['b1'],name='fc_1')
     layer_1 = tf.nn.relu(layer_1,name='relu_1')

     layer_2 = tf.add(tf.matmul(layer_1,weights['h2']),biases['b2'],name='fc_2')
     layer_2 = tf.nn.relu(layer_2,name='relu_2')

     out_layer = tf.add(tf.matmul(layer_2,weights['out']),biases['out'],name='fc_3')

     return out_layer

pred = multilayer_perceptron(X, weights, biases)

learning_rate = 0.001
loss_all = tf.nn.softmax_cross_entropy_with_logits(labels=Y,logits=pred,name='cross_entropy')
loss = tf.reduce_mean(loss_all,name='avg_loss')
optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(loss)

init =tf.global_variables_initializer()

training_epochs=15

batch_size=128
display_step =1

with tf.Session() as sess:
    sess.run(init)
    writer =tf.summary.FileWriter('./graphs/MLP_dnn',sess.graph)

    for epoch in range(training_epochs):
        avg_loss =0
        total_batch = int(mnist.train.num_examples/batch_size)

        for i in range(total_batch):
            batch_x, batch_y =mnist.train.next_batch(batch_size)
            _, l =sess.run([optimizer,loss],feed_dict={X: batch_x, Y: batch_y})

            avg_loss += l/total_batch

        if epoch%display_step==0:
            print ('Epoch:', '%04d'% (epoch+1),'cost=', "{:.9f}".format(avg_loss))

    print ('optimizer finished')

    correct_prediction = tf.equal(tf.argmax(pred,1),tf.argmax(Y,1))

    accuracy = tf.reduce_mean(tf.cast(correct_prediction,'float'))
    print ('accuracy:',accuracy.eval({X:mnist.test.images, Y:mnist.test.labels}))
    writer.close()