basicCNN.py 4.66 KB
Newer Older
AF3's avatar
AF3 committed
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
### Ali Farahmand
### CS 488 Senior Capstone


#imports
import numpy as np
import pandas as pd
from keras.preprocessing.image import ImageDataGenerator
from keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout
from keras.models import Sequential
from keras import optimizers
from tensorflow.keras.callbacks import ReduceLROnPlateau, ModelCheckpoint
import matplotlib
import matplotlib.pyplot as plt
import matplotlib.image as mpimg

#commonly used variables
batch_size = 10
image_size = (96,96)
epochs = 10


#loading csv into dataframe
labels = pd.read_csv("histopathologic-cancer-detection/train_labels.csv")


#Creating the training and valiation sets
labels_neg = labels[labels['label'] == 0].sample(80000, random_state=3)
labels_pos = labels[labels['label'] == 1].sample(80000, random_state=3)

val_neg = labels_neg[:8000]
val_pos = labels_pos[:8000]

labels_neg = pd.concat([labels_neg, val_neg]).drop_duplicates(keep=False)
labels_pos = pd.concat([labels_pos, val_pos]).drop_duplicates(keep=False)

train_labels = pd.concat([labels_neg, labels_pos])
val_labels = pd.concat([val_neg, val_pos])

#adding image format .tif to the end of each id
def append_ext(fn):
    return fn+".tif"
train_labels["id"] = train_labels["id"].apply(append_ext)
val_labels["id"] = val_labels["id"].apply(append_ext)


#datagenerators and image augmentation
train_datagen = ImageDataGenerator(rescale=1./255, zoom_range=0.3, rotation_range=50,
									width_shift_range=0.2, height_shift_range=0.2, shear_range=0.2, 
									horizontal_flip=True, fill_mode='nearest')

val_datagen = ImageDataGenerator(rescale=1./255)


train_generator = train_datagen.flow_from_dataframe(dataframe=train_labels, directory="histopathologic-cancer-detection/train", x_col="id", y_col="label", class_mode="raw", target_size=image_size, batch_size=batch_size)
val_generator = val_datagen.flow_from_dataframe(dataframe=val_labels, directory="histopathologic-cancer-detection/train", x_col="id", y_col="label", class_mode="raw", target_size=image_size, batch_size=batch_size)

#Creating the model
model = Sequential()

model.add(Conv2D(32, kernel_size=(3, 3), activation = 'relu', input_shape = (96, 96, 3)))
model.add(Conv2D(32, kernel_size=(3, 3), activation = 'relu'))
model.add(Conv2D(32, kernel_size=(3, 3), activation = 'relu'))
model.add(MaxPooling2D(pool_size = (2,2))) 
model.add(Dropout(0.3))

model.add(Conv2D(64, kernel_size=(3, 3), activation ='relu'))
model.add(Conv2D(64, kernel_size=(3, 3), activation ='relu'))
model.add(Conv2D(64, kernel_size=(3, 3), activation ='relu'))
model.add(MaxPooling2D(pool_size = (2,2)))
model.add(Dropout(0.3))

model.add(Conv2D(128, kernel_size=(3, 3), activation ='relu'))
model.add(Conv2D(128, kernel_size=(3, 3), activation ='relu'))
model.add(Conv2D(128, kernel_size=(3, 3), activation ='relu'))
model.add(MaxPooling2D(pool_size = (2,2)))
model.add(Dropout(0.3))

model.add(Flatten())
80
81
model.add(Dense(512, activation = "relu"))
model.add(Dropout(0.5))
AF3's avatar
AF3 committed
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
model.add(Dense(1, activation = "sigmoid"))

model.compile(loss='binary_crossentropy',
			  optimizer=optimizers.Adam(lr=1e-4),
              metrics=['accuracy'])

model.summary()


#saving the model at best val accuracy
filepath = "model.h5"
checkpoint = ModelCheckpoint(filepath, monitor='val_accuracy', verbose=1, 
                             save_best_only=True, mode='max')
#reducing the learning rate if val accuracy drops
reduce_lr = ReduceLROnPlateau(monitor='val_accuracy', factor=0.5, patience=2, 
                                   verbose=1, mode='max', min_lr=0.00001)
                                                         
callbacks_list = [checkpoint, reduce_lr]

#training
history = model.fit(train_generator, batch_size = batch_size, 
                    validation_data=val_generator,
                    validation_steps=50,
                    epochs=epochs, verbose=1,
                    callbacks = callbacks_list)

108
#graphing the accuracy and loss values
AF3's avatar
AF3 committed
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 4))
t = f.suptitle('Basic CNN Performance', fontsize=12)
f.subplots_adjust(top=0.85, wspace=0.3)

epoch_list = list(range(1,epochs+1))
ax1.plot(epoch_list, history.history['accuracy'], label='Train Accuracy')
ax1.plot(epoch_list, history.history['val_accuracy'], label='Validation Accuracy')
ax1.set_xticks(np.arange(0, epochs+1, 5))
ax1.set_ylabel('Accuracy Value')
ax1.set_xlabel('Epoch')
ax1.set_title('Accuracy')
l1 = ax1.legend(loc="best")

ax2.plot(epoch_list, history.history['loss'], label='Train Loss')
ax2.plot(epoch_list, history.history['val_loss'], label='Validation Loss')
ax2.set_xticks(np.arange(0, epochs+1, 5))
ax2.set_ylabel('Loss Value')
ax2.set_xlabel('Epoch')
ax2.set_title('Loss')
l2 = ax2.legend(loc="best")
plt.savefig('plotmain.png')