-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathbagging _cifar10.py
173 lines (148 loc) · 7.29 KB
/
bagging _cifar10.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
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
import cifar10
import cifar10_input
import tensorflow as tf
import numpy as np
import time
from collections import Counter
max_steps =500
batch_size =128
data_dir = '/tmp/cifar10_data/cifar-10-batches-bin'
cifar10.maybe_download_and_extract()
images_train,labels_train = cifar10_input.distorted_inputs(data_dir=data_dir,batch_size=batch_size)
images_test ,labels_test = cifar10_input.inputs(eval_data=True, data_dir= data_dir,batch_size=batch_size)
image_holder = tf.placeholder(tf.float32,[batch_size,24,24,3])
label_holder = tf.placeholder(tf.int32,[batch_size])
def constant_variable_biases(value,shape,is_var=1):
if is_var == 1:
var = tf.Variable(tf.constant(value,shape=shape))
else:
var = tf.constant(value=value,shape=shape)
return var
def constant_variable_weight(shape,stddev,is_var=1):
if is_var == 1:#等于1取变量，等于0取常量固定
var = tf.Variable(tf.truncated_normal(shape,stddev=stddev))
else:
var = tf.constant(np.array(np.random.normal(0,stddev,shape),np.float32))
return var
def conv2d(x,W,b,strides=1):
x = tf.nn.conv2d(x,W,strides=[1,strides,strides,1],padding='SAME')
x = tf.nn.bias_add(x,b)
return x
def maxpool2d(x,k=2):
return tf.nn.max_pool(x,ksize=[1,k+1,k+1,1],strides=[1,k,k,1],padding='SAME')
def conv_net(image_holder,weights,biases):
conv1 = tf.nn.relu(conv2d(image_holder,weights['wc1'],biases['bc1']))
pool1 = maxpool2d(conv1)
#
# conv2 = tf.nn.relu(conv2d(pool1,weights['wc2'],biases['bc2']))
# pool2 = maxpool2d(conv2)
fc1 = tf.reshape(pool1, [-1, weights['out'].get_shape().as_list()[0]])
# fc1 = tf.add(tf.matmul(fc1, weights['wd1']), biases['bd1'])
# fc1 = tf.nn.relu(fc1)
# fc1 = tf.nn.dropout(fc1, dropout)
# fc2 = tf.add(tf.matmul(fc1, weights['wd2']), biases['bd2'])
# fc2 = tf.nn.relu(fc2)
out = tf.add(tf.matmul(fc1, weights['out']), biases['out'])
return out
def create_graph():
weights = {
'wc1': constant_variable_weight([5,5,3,64],stddev=0.1,is_var=0),
# 'wc2': constant_variable_weight([5,5,32,64],stddev=5e-2),
# 'wd1': constant_variable_weight([12*12*32,1024],stddev=0.04),
# # 'wd2': constant_with_weight_loss([384,192],stddev=0.04),
'out': constant_variable_weight([12*12*64,10],stddev=1/192.0),
}
biases = {
'bc1': constant_variable_biases(0.0,shape=[64],is_var=0),
# 'bc2': constant_variable_biases(0.1,shape=[64]),
# 'bd1': constant_variable_biases(0.1,shape=[1024]),
# # 'bd2': constant_variable_biases(0.1,shape=[192]),
'out': constant_variable_biases(0.0,shape=[10])
}
logits = conv_net(image_holder,weights,biases)
pred = tf.argmax(tf.nn.softmax(logits),1)
cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=logits, labels=label_holder,
name='cross_entropy_per_example')
loss = tf.reduce_mean(cross_entropy, name='cross_entropy')
train_op = tf.train.AdamOptimizer(1e-3).minimize(loss)
top_k_op = tf.nn.in_top_k(logits,label_holder,1)
return pred,weights,biases,loss,train_op,top_k_op
pred,weights,biases,loss,train_op,top_k_op =create_graph()
pred2,weights2,biases2,loss2,train_op2,top_k_op2 =create_graph()
pred3,weights3,biases3,loss3,train_op3,top_k_op3 =create_graph()
pred4,weights4,biases4,loss4,train_op4,top_k_op4 =create_graph()
pred5,weights5,biases5,loss5,train_op5,top_k_op5 =create_graph()
sess = tf.InteractiveSession()
tf.global_variables_initializer().run()
tf.train.start_queue_runners()
#训练开始
train_time = time.time()
for step in range(max_steps):
start_time = time.time()
image_batch,label_batch = sess.run([images_train,labels_train])
image_batch2,label_batch2 = sess.run([images_train,labels_train])
image_batch3,label_batch3 = sess.run([images_train,labels_train])
image_batch4,label_batch4 = sess.run([images_train,labels_train])
image_batch5,label_batch5 = sess.run([images_train,labels_train])
_,loss_value =sess.run([train_op,loss],feed_dict={image_holder:image_batch,label_holder:label_batch})
_,loss_value2 =sess.run([train_op2,loss2],feed_dict={image_holder:image_batch2,label_holder:label_batch2})
_,loss_value3 =sess.run([train_op3,loss3],feed_dict={image_holder:image_batch3,label_holder:label_batch3})
_,loss_value4 =sess.run([train_op4,loss4],feed_dict={image_holder:image_batch4,label_holder:label_batch4})
_,loss_value5 =sess.run([train_op5,loss5],feed_dict={image_holder:image_batch5,label_holder:label_batch5})
duration = time.time() - start_time
if step %100 == 0:
examples_per_sec = batch_size/duration
sec_per_batch = float(duration)
format_str = ('step %d,loss=%.2f(%.1f examples/sec; %.3f sec/batch')
print(format_str % (step,loss_value,examples_per_sec,sec_per_batch))
print(format_str % (step,loss_value2,examples_per_sec,sec_per_batch))
print(format_str % (step,loss_value3,examples_per_sec,sec_per_batch))
print(format_str % (step,loss_value4,examples_per_sec,sec_per_batch))
print(format_str % (step,loss_value5,examples_per_sec,sec_per_batch))
print('train_time: ',time.time() - train_time)
#测试开始
num_examples =10000
import math
num_iter = int(math.ceil(num_examples/batch_size))
true_count =[0,0,0,0,0,0]
total_sample_count = num_iter *batch_size
step =0
while step < num_iter:
image_batch,label_batch = sess.run([images_test,labels_test])
predictions,acc = sess.run([pred,top_k_op],feed_dict={image_holder:image_batch,label_holder:label_batch})
predictions2,acc2 = sess.run([pred2,top_k_op2],feed_dict={image_holder:image_batch,label_holder:label_batch})
predictions3,acc3 = sess.run([pred3,top_k_op3],feed_dict={image_holder:image_batch,label_holder:label_batch})
predictions4,acc4 = sess.run([pred4,top_k_op4],feed_dict={image_holder:image_batch,label_holder:label_batch})
predictions5,acc5 = sess.run([pred5,top_k_op5],feed_dict={image_holder:image_batch,label_holder:label_batch})
temppred =[]
temppred.append(predictions)
temppred.append(predictions2)
temppred.append(predictions3)
temppred.append(predictions4)
temppred.append(predictions5)
#获得投票结果
temppred = np.array(temppred)
temp =np.zeros(batch_size)
for j in range(batch_size):
word_counts = Counter(temppred[:,j])
temp[j] = word_counts.most_common(1)[0][0]
acc6 = np.equal(temp,label_batch)
true_count[0]+=np.sum(acc)
true_count[1]+=np.sum(acc2)
true_count[2]+=np.sum(acc3)
true_count[3]+=np.sum(acc4)
true_count[4]+=np.sum(acc5)
true_count[5]+=np.sum(acc6)
step +=1
precision = true_count[0]/total_sample_count
precision2 = true_count[1]/total_sample_count
precision3 = true_count[2]/total_sample_count
precision4 = true_count[3]/total_sample_count
precision5 = true_count[4]/total_sample_count
end_pred = true_count[5]/total_sample_count
print('precision @ 1 =%.5f' % precision)
print('precision @ 1 =%.5f' % precision2)
print('precision @ 1 =%.5f' % precision3)
print('precision @ 1 =%.5f' % precision4)
print('precision @ 1 =%.5f' % precision5)
print('precision @ 1 =%.5f' % end_pred)