小白学Tensorflow之Logistic回归

利用Tensorflow实现Logistic回归

第一，我们先来设计两个函数，使得在后续的程序中不用重复编写相同的代码。

def init_weights(shape):
return tf.Variable(tf.random_normal(shape, stddev = 0.01))

def model(X, w):
return tf.matmul(X, w)

第二，我们带入mnist的数据集，具体方法可以参考官网。

# 导入数据
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
trX, trY, teX, teY = mnist.train.images, mnist.train.labels, mnist.test.images, mnist.test.labels

第三，构建损失函数，我们采用softmax和交叉熵来训练模型

# 构建损失函数，我们采用softmax和交叉熵来训练模型
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
learning_rate = 0.01
train_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)

完整代码如下：

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import numpy as np
import tensorflow as tf
import input_data

def init_weights(shape):
return tf.Variable(tf.random_normal(shape, stddev = 0.01))

def model(X, w):
return tf.matmul(X, w)

# 导入数据
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
trX, trY, teX, teY = mnist.train.images, mnist.train.labels, mnist.test.images, mnist.test.labels

# 设置占位符
X = tf.placeholder("float", [None, 784])
Y = tf.placeholder("float", [None, 10])

# 初始化权重
w = init_weights([784, 10])

# 构建模型
py_x = model(X, w)

# 构建损失函数，我们采用softmax和交叉熵来训练模型
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
learning_rate = 0.01
train_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)
predict_op = tf.argmax(py_x, 1)

with tf.Session() as sess:

init = tf.initialize_all_variables()
sess.run(init)

for i in xrange(100):
for start, end in zip(range(0, len(trX), 128), range(128, len(trX), 128)):
sess.run(train_op, feed_dict = {X: trX[start:end], Y: trY[start:end]})
print i, np.mean(np.argmax(teY, axis = 1) == sess.run(predict_op, feed_dict = {X: teX, Y: teY}))

简书同步更新：http://www.jianshu.com/p/f51f0ca4278c

84cb