Tensorflow-Mnist-MLP

Tensorflow Mnist /MLP/BP/CNN

Tnesorflow 1.7 用3层隐藏层实现手写体识别

import tensorflow as tf import numpy as np import matplotlib.pyplot as plt import os

解决CPU不匹配的 忽略码问题 ‘1’，‘2’，‘3’

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' os.environ["CUDA_VISIBLE_DEVICES"] = "0"

加载MNIST数据集

从官网直接下载

从本地加载数据

mnist_folder="D:\datamist\MNIST" from tensorflow.examples.tutorials.mnist import input_data mnist = input_data.read_data_sets(mnist_folder,one_hot=True)

定义模型参数

n_input = 784 n_labels = 10 n_hidden1 = 128 n_hidden2 = 256

定义模型 预测初始化的占位符 x y

X = tf.placeholder(tf.float32,[None,n_input]) Y = tf.placeholder(tf.float32,[None,n_labels])

构建模型

权值 偏置 数据字典 键值对方式 去获取权值 偏置参数

weights ={ 'h1': tf.Variable(tf.truncated_normal([n_input,n_hidden1],stddev=0.1)), 'h2': tf.Variable(tf.truncated_normal([n_hidden1,n_hidden2],stddev=0.1)), 'out':tf.Variable(tf.truncated_normal([n_hidden2,n_labels],stddev=0.1)) } biases= { 'h1':tf.Variable(tf.random_normal([n_hidden1])), 'h2':tf.Variable(tf.random_normal([n_hidden2])), 'out':tf.Variable(tf.random_normal([n_labels])) }

def mlpmodel(X,weights,biases): layer1 = tf.nn.relu(tf.add(tf.matmul(X,weights['h1']),biases['h1'])) layer2 = tf.nn.relu(tf.add(tf.matmul(layer1,weights['h2']),biases['h2'])) outlayer = tf.add(tf.matmul(layer2,weights['out']),biases['out']) return outlayer

前向计算网络模型输出

pred = mlpmodel(X,weights,biases)

反向传播 交叉熵

cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(logits=pred,labels=Y))

优化器

optimizer = tf.train.AdamOptimizer(learning_rate=0.001).minimize(cost)

模型预测 模型训练

准确率 最大似然估计

correct_prediction = tf.equal(tf.argmax(pred,1),tf.argmax(Y,1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32))

epoches =1000 display_step=100 batch_size = 50 avg_loss = 0.0

模型训练

with tf.Session() as sess: sess.run(tf.global_variables_initializer()) for epoch in range(epoches): 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) sess.run(optimizer,feed_dict={X: batch_x,Y:batch_y}) loss,acc = sess.run([cost,accuracy],feed_dict={X:mnist.train.images,Y:mnist.train.labels}) avg_loss += loss/total_batch if(epoch+1) % display_step ==0: print('epoch:',epoch+1,'cost=','{:.6f}'.format(avg_loss),'accuracy=','{:.6f}'.format(acc)) print("train finished")