-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathneural_network.py
227 lines (174 loc) · 9.26 KB
/
neural_network.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
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
import numpy as np
import cv2
import skimage.color
import skimage.filters
import skimage.io
from sklearn.model_selection import train_test_split
from sklearn import preprocessing
import pickle
from sklearn.base import BaseEstimator, ClassifierMixin
from sklearn.utils import check_random_state
from sklearn.preprocessing import LabelEncoder
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D, BatchNormalization, Conv2DTranspose, Activation,\
Concatenate
from keras.losses import sparse_categorical_crossentropy
from keras.optimizers import adadelta, adam
from keras.models import load_model
from keras.callbacks import ReduceLROnPlateau
from preprocess_data import load_data
image_size = 256
X_Train_Mass = np.load('X_mass_train_' + str(image_size) + '.npy')
X_Train_Calc = np.load('X_calc_train_' + str(image_size) + '.npy')
X_Test_Calc = np.load('X_calc_test_' + str(image_size) + '.npy')
X_Test_Mass = np.load('X_mass_test_' + str(image_size) + '.npy')
Y_Train_Mass = np.load('Y_mass_train.npy')
Y_Train_Calc = np.load('Y_calc_train.npy')
Y_Test_Calc = np.load('Y_calc_test.npy')
Y_Test_Mass = np.load('Y_mass_test.npy')
print("Mass Train Benign: " + str(np.count_nonzero(Y_Train_Mass == 0)))
print("Mass Train Malignant: " + str(np.count_nonzero(Y_Train_Mass == 1)))
print("Calc Train Benign: " + str(np.count_nonzero(Y_Train_Calc == 0)))
print("Calc Train Malignant: " + str(np.count_nonzero(Y_Train_Calc == 1)))
print("Mass Test Benign: " + str(np.count_nonzero(Y_Test_Mass == 0)))
print("Mass Test Malignant: " + str(np.count_nonzero(Y_Test_Mass == 1)))
print("Calc Test Benign: " + str(np.count_nonzero(Y_Test_Calc == 0)))
print("Calc Test Malignant: " + str(np.count_nonzero(Y_Test_Calc == 1)))
print("X_Train_Mass.shape = " + str(X_Train_Mass.shape))
X_Train_Mass_count = X_Train_Mass.shape[2]
print("X_Test_Mass.shape = " + str(X_Test_Mass.shape))
X_Val_Mass = X_Test_Mass[:189, :, :, :]
Y_Val_Mass = Y_Test_Mass[0: 189]
X_Test_Mass = X_Test_Mass[189:, :, :, :]
Y_Test_Mass = Y_Test_Mass[189:]
print(X_Train_Mass.shape)
print(Y_Train_Mass.shape)
print(X_Test_Mass.shape)
print(Y_Test_Mass.shape)
print(X_Val_Mass.shape)
print(Y_Val_Mass.shape)
batches_per_epoch = int(1318 / 32)
print("Steps per epoch: ", batches_per_epoch)
lr_decay = (1./0.80 - 1) / batches_per_epoch
epoch_count = 30
model_cnn = Sequential()
#conv layer 1
model_cnn.add(Conv2D(32, kernel_size=(3, 3), strides=(2, 2), padding='same', input_shape=(image_size, image_size, 3)))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#conv layer 2
model_cnn.add(Conv2D(32, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#conv layer 3
model_cnn.add(Conv2D(32, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#max pool 1
model_cnn.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
model_cnn.add(Dropout(0.5))
#conv layer 4
model_cnn.add(Conv2D(64, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#conv layer 5
model_cnn.add(Conv2D(64, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#max pool 2
model_cnn.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
model_cnn.add(Dropout(0.5))
#conv layer 6
model_cnn.add(Conv2D(128, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#conv layer 7
model_cnn.add(Conv2D(128, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#max pool 3
model_cnn.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
#conv layer 8
model_cnn.add(Conv2D(256, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
#conv layer 9
#model_cnn.add(Conv2D(256, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#max pool 4
model_cnn.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
#conv layer 10
model_cnn.add(Conv2D(512, kernel_size=(3, 3), strides=(1, 1), padding='same'))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
##conv layer 11
#model_cnn.add(Conv2D(512, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#max pool 5
model_cnn.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))
##conv layer 12
#model_cnn.add(Conv2D(512, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#conv layer 13
#model_cnn.add(Conv2D(512, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#max pool 6
#model_cnn.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same', name='pool6'))
# transpose to 256
#unpool1 = Conv2DTranspose(256, kernel_size=(8, 8), strides=(4, 4), activation='elu')
#model_cnn.add(unpool1)
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#unpool1 = Conv2D(256, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same')
#model_cnn.add(unpool1)
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#
#unpool2 = Conv2D(128, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same')
#model_cnn.add(unpool2)
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#unpool2 = Conv2D(256, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same')
#model_cnn.add(unpool2)
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
##
#merged = Concatenate([unpool1, unpool2])
#model_cnn.add(Conv2DTranspose(256, kernel_size=(4, 4), strides=(2, 2), activation='elu'))
#model_cnn.add(Conv2D(128, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.#add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#model_cnn.add(Conv2D(56, kernel_size=(1, 1), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#model_cnn.add(Concatenate(axis=-1))
#model_cnn.add(Conv2DTranspose(128, kernel_size=(3, 3), strides=(1, 1), activation='elu'))
#model_cnn.add(Conv2D(64, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#
#model_cnn.add(Conv2D(32, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#
#model_cnn.add(Conv2D(16, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Flatten())
model_cnn.add(Dropout(0.5))
model_cnn.add(Dense(2048, activation=None))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
model_cnn.add(Dropout(0.5))
model_cnn.add(Dense(2048, activation=None))
model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
model_cnn.add(Activation('relu'))
model_cnn.add(Dropout(0.5))
model_cnn.add(Dense(2, activation=None))
#model_cnn.add(Conv2D(2, kernel_size=(3, 3), strides=(1, 1), activation='relu', padding='same'))
#model_cnn.add(BatchNormalization(axis=-1, momentum=0.99, epsilon=1e-8, center=True, scale=True))
#model_cnn.add(Conv2DTranspose(16, kernel_size=(4, 4), strides=(2, 2), activation='elu'))
#
#model_cnn.add(Conv2DTranspose(2, kernel_size=(3, 3), strides=(1, 1), activation='softmax'))
opt = adam(lr=0.001, decay=lr_decay)
model_cnn.compile(optimizer=opt, loss=sparse_categorical_crossentropy, metrics=['accuracy'])
model_cnn.fit(X_Train_Mass, Y_Train_Mass,
batch_size=32,
epochs=epoch_count,
verbose=1,
validation_data=(X_Val_Mass, Y_Val_Mass))
model_cnn.save(str(image_size) + "px_" + str(epoch_count) + "epoch" + '_mass' + ".h5")
score = model_cnn.evaluate(X_Test_Mass, Y_Test_Mass, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])