-
Notifications
You must be signed in to change notification settings - Fork 1
/
cnn_train.py
85 lines (66 loc) · 3.09 KB
/
cnn_train.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
# coding:utf-8
import tensorflow as tf
import numpy as np
import data_input
def weight_variable(shape):
initial = tf.truncated_normal(shape, stddev=0.1) # 变量的初始值为截断正太分布
return tf.Variable(initial)
def bias_variable(shape):
initial = tf.constant(0.1, shape=shape)
return tf.Variable(initial)
def conv2d(x, W):
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
sess = tf.InteractiveSession()
x = tf.placeholder(tf.float32, [None, 80],name='x')
x_image = tf.reshape(x, [-1,8,10,1]) #将输入按照 conv2d中input的格式来reshape,reshape
"""
# 第一层
"""
W_conv1 = weight_variable([3, 3, 1, 12])
b_conv1 = bias_variable([12])
h_conv1 = tf.nn.elu(conv2d(x_image, W_conv1) + b_conv1)
"""
# 第二层
"""
W_conv2 = weight_variable([3, 3, 12, 24])
b_conv2 = bias_variable([24])
h_conv2 = tf.nn.elu(conv2d(h_conv1, W_conv2) + b_conv2)
# 第三层 是个全连接层,输入维数16*5*64, 输出维数为1024
W_fc1 = weight_variable([8 * 10 * 24, 1024])
b_fc1 = bias_variable([1024])
h_conv2_flat = tf.reshape(h_conv2, [-1, 8*10*24])
h_fc1 = tf.nn.elu(tf.matmul(h_conv2_flat, W_fc1) + b_fc1)
keep_prob = tf.placeholder(tf.float32,name='keep_prob') # 这里使用了drop out,即随机安排一些cell输出值为0,可以防止过拟合
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
# 第四层,输入1024维,输出3维,也就是具体的0~2分类
W_fc2 = weight_variable([1024, 3])
b_fc2 = bias_variable([3])
y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2) + b_fc2) # 使用softmax作为多分类激活函数
y_ = tf.placeholder(tf.float32, [None, 3],name='y_')
#y_conv+1e-10 特别重要
cross_entropy = tf.reduce_mean(-tf.reduce_sum(y_ * tf.log(y_conv+1e-10), reduction_indices=[1]))# 损失函数,交叉熵
tf.summary.scalar("loss", cross_entropy)
train_step = tf.train.AdamOptimizer(10e-6).minimize(cross_entropy) # 使用adam优化 过大会造成梯度爆炸
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1)) # 计算准确度
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32),name='accuracy')
tf.summary.scalar('Accuracy', accuracy)
merged_summary = tf.summary.merge_all()
writer = tf.summary.FileWriter("logs/", sess.graph)
sess.run(tf.global_variables_initializer()) # 变量初始化
saver = tf.train.Saver() # 声明tf.train.Saver类用于保存模型
for i in range(8500):
batch = data_input.InputClass().next_batch(50)
if i%50 == 0:
print('data:',batch[0].shape,'label:',batch[1].shape)
train_accuracy = accuracy.eval(feed_dict={
x:batch[0], y_: batch[1], keep_prob: 1.0})
result = sess.run(merged_summary,
feed_dict={
x: batch[0], y_: batch[1], keep_prob: 1.0})
writer.add_summary(result, i)
print("step %d, training accuracy %g"%(i, train_accuracy))
train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5})
saver.save(sess, 'save/model.ckpt')
testData = data_input.InputClass().test_data()
print("test accuracy %g"%accuracy.eval(feed_dict={
x: testData[0], y_: testData[1], keep_prob: 1.0}))