tensorflow入门（二）

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

#使用numpy生成200个随机点
x_data = np.linspace(-0.5,0.5,200)[:,np.newaxis]
noise = np.random.normal(0,0.02,x_data.shape)
y_data = np.square(x_data) + noise

#定义两个placeholder
x = tf.placeholder(tf.float32,[None,1])
y = tf.placeholder(tf.float32,[None,1])

#定义神经网络中间层
Weights_L1 = tf.Variable(tf.random_normal([1,10]))
biases_L1 = tf.Variable(tf.zeros([1,10]))
Wx_plus_b_L1 = tf.matmul(x,Weights_L1) + biases_L1
L1 = tf.nn.tanh(Wx_plus_b_L1)

#定义神经网络输出层
Weights_L2 = tf.Variable(tf.random_normal([10,1]))
biases_L2 = tf.Variable(tf.zeros([1,1]))
Wx_plus_b_L2 = tf.matmul(L1,Weights_L2) + biases_L2
prediction = tf.nn.tanh(Wx_plus_b_L2)

#二次代价函数
loss = tf.reduce_mean(tf.square(y - prediction))
#使用梯度下降法
train_step = tf.train.GradientDescentOptimizer(0.1).minimize(loss)

with tf.Session() as sess:
    #变量初始化
    sess.run(tf.global_variables_initializer())
    for _ in range(2000):
        sess.run(train_step,feed_dict={x:x_data,y:y_data})


    #获得预测值
    prediction_value = sess.run(prediction,feed_dict={x:x_data})
    #画图
    plt.figure()
    plt.scatter(x_data,y_data)
    plt.plot(x_data,prediction_value,'r-',lw = 5)
    plt.show()

得到结果：

——

import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

#载入数据集
mnist = input_data.read_data_sets("MNIST_data",one_hot=True)

#每个批次的大小
batch_size = 100
#计算一共有多少个批次
n_batch = mnist.train.num_examples // batch_size

#定义两个placeholder
x = tf.placeholder(tf.float32,[None,784]) #图片
y = tf.placeholder(tf.float32,[None,10]) #标签

#创建一个简单的神经网络
w = tf.Variable(tf.zeros([784,10]))
b = tf.Variable(tf.zeros([10]))
prediction = tf.nn.softmax(tf.matmul(x,w)+b)

#二次代价函数
loss = tf.reduce_mean(tf.square(y-prediction))
#梯度下降算法
train_step = tf.train.GradientDescentOptimizer(0.2).minimize(loss)

#初始化变量
init = tf.global_variables_initializer()

#结果存放在一个bool类型的列表中,argmax()返回一维张量中最大值所在的位置，equal函数判断两者是否相等
correct_prediction = tf.equal(tf.argmax(y,1),tf.argmax(prediction,1))
#求准确率,cast：将bool型转换为0，1然后求平均正好算到是准确率
accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32))

with tf.Session() as sess:
    sess.run(init)
    for epoch in range(21):
        for batch in range(batch_size):#next_batch:不断获取下一组数据
            batch_xs,batch_ys = mnist.train.next_batch(batch_size)
            sess.run(train_step,feed_dict={x:batch_xs,y:batch_ys})

        acc = sess.run(accuracy,feed_dict={x:mnist.test.images,y:mnist.test.labels})
        print("Iter "+str(epoch)+",Testing Accuracy "+str(acc))

训练结果：

Iter 0,Testing Accuracy 0.657
Iter 1,Testing Accuracy 0.7063
Iter 2,Testing Accuracy 0.7458
Iter 3,Testing Accuracy 0.7876
Iter 4,Testing Accuracy 0.8203
Iter 5,Testing Accuracy 0.8406
Iter 6,Testing Accuracy 0.8492
Iter 7,Testing Accuracy 0.8586
Iter 8,Testing Accuracy 0.8642
Iter 9,Testing Accuracy 0.8678
Iter 10,Testing Accuracy 0.8709
Iter 11,Testing Accuracy 0.8736
Iter 12,Testing Accuracy 0.8751
Iter 13,Testing Accuracy 0.8784
Iter 14,Testing Accuracy 0.88
Iter 15,Testing Accuracy 0.8816
Iter 16,Testing Accuracy 0.883
Iter 17,Testing Accuracy 0.8841
Iter 18,Testing Accuracy 0.8847
Iter 19,Testing Accuracy 0.8855
Iter 20,Testing Accuracy 0.8878

可见，通过多次的训练，识别的准确值不断提高

——