-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata_cvs2Train_CNN.py
143 lines (135 loc) · 5.33 KB
/
data_cvs2Train_CNN.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
from tkinter.tix import Y_REGION
import pandas as pd
import numpy as np
#load data
#train
train_data = pd.read_csv("tr_data.csv")
#train_data =train_data[~np.isnan(train_data)]
train_data = train_data.drop(index=0)
train_data = train_data.values.reshape(10751, 120,1,1)
print("train_data_shape:",train_data.shape)
#print("len to train_data",len(train_data))
X_train =train_data
#test
test_data = pd.read_csv("tst_data.csv")
test_data = test_data.drop(index=0)
test_data = test_data.values.reshape(2575, 120,1,1)
#test_data =test_data[~np.isnan(test_data)]
print("test_data_shape:",test_data.shape)
#print("len to test_data",len(test_data))
X_test =test_data
#train_label
train_lab = pd.read_csv("tr_label.csv")
train_lab = train_lab.values.reshape(-1,1)
print(train_lab.shape)
#test_label
test_lab = pd.read_csv("tst_label.csv")
test_lab = test_lab.values.reshape(-1,1)
print(test_lab.shape)
############ shuffle ###########
index1= [i for i in range(len(X_train))]
np.random.shuffle(index1)
X_train=X_train[index1]
train_lab = train_lab[index1]
index2= [i for i in range(len(X_test))]
np.random.shuffle(index2)
X_test=X_test[index2]
test_lab =test_lab[index2]
####################################
##############std###################
X_train -= X_train.mean(axis=0)
X_train /= X_train.std(axis=0)
X_test -= X_test.mean(axis=0)
X_test /= X_test.std(axis=0)
####################
#load tensorflow/keras
from tensorflow import keras
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Dense, GlobalAveragePooling2D
from keras.utils import np_utils
import matplotlib.pyplot as plt
from sklearn.metrics import confusion_matrix, roc_curve, auc
import sklearn.metrics as metrics
#y_trainOneHot= pd.get_dummies(train_lab)
#y_testOneHot= pd.get_dummies(test_lab)
y_trainOneHot = np_utils.to_categorical(train_lab)
y_testOneHot = np_utils.to_categorical(test_lab)
#print(y_testOneHot[0].shape)
input_shape = (120,1,1)
def model_by_self(input_shape=input_shape):
from tensorflow.keras.layers import PReLU,LeakyReLU
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten, Conv2D, MaxPooling2D, ZeroPadding2D,BatchNormalization
model = Sequential()
model.add(Conv2D(filters=8, kernel_size=(3, 3), padding='same', input_shape=(120,1,1),activation = 'relu'))
#model.add(LeakyReLU(alpha=.001)) # add an advanced activation
model.add(BatchNormalization())
#model.add(PReLU())
#model.add(Dropout(rate=0.25))
model.add(MaxPooling2D(pool_size=(2, 1)))
model.add(Conv2D(filters=16, kernel_size=(3, 3), padding='same',activation = 'relu'))
model.add(BatchNormalization())
#model.add(LeakyReLU(alpha=.001)) # add an advanced activation
#model.add(PReLU())
#model.add(Dropout(rate=0.2))
model.add(MaxPooling2D(pool_size=(2, 1)))
model.add(Conv2D(filters=32, kernel_size=(3, 3), padding='same',activation = 'relu'))
model.add(BatchNormalization())
#model.add(PReLU())
model.add(MaxPooling2D(pool_size=(2, 1)))
#model.add(Dropout(rate=0.25))
model.add(Flatten())
model.add(Dropout(rate=0.5))
model.add(Dense(128, activation='relu'))
model.add(LeakyReLU(alpha=.001)) # add an advanced activation
#model.add(PReLU())
model.add(Dropout(rate=0.25))
#model.add(Dense(16, activation='relu'))
#model.add(Dropout(rate=0.2))
#model.add(Dense(256, activation='relu'))
#model.add(Dropout(rate=0.25))
#model.add(Dense(64, activation='relu'))
#model.add(Dropout(rate=0.25))
model.add(Dense(15, activation='softmax'))
return model
model = model_by_self(input_shape=input_shape)
model.compile(loss='categorical_crossentropy', optimizer=keras.optimizers.Adam(lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=False), metrics=['accuracy'])
#summary
model.summary()
epochs = 30
batch_size = 64
steps_per_epoch = int(len(X_train)*2/batch_size)
samples_per_epoch=(len(X_train)*2)
##########Tensorboard#####
#from keras.callbacks import TensorBoard
#from keras.callbacks import EarlyStopping, ModelCheckpoint
#import tensorflow as tf
#import datetime
#log_dir = "logs" #+ datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
#tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1,write_graph=True, write_images=True)
############################
history = model.fit(x=X_train, y=y_trainOneHot, batch_size=8, epochs=100, verbose=1, validation_split=0.2)
scores = model.evaluate(X_test,y_testOneHot,verbose=1)
print("scores:",scores[1])
prediction= model.predict_classes(X_test)
cnf_matrix = metrics.confusion_matrix(test_lab, prediction)
############acc / loss###########
def plt_loss(history):
fig = plt.figure(figsize=(15, 5))
ax1 = fig.add_subplot(1, 2, 1)
ax1.plot(history.history['accuracy'])
ax1.plot(history.history['val_accuracy'])
ax1.set_title('model accuracy')
ax1.set_ylabel('accuracy')
ax1.set_xlabel('epoch')
ax1.legend(['train', 'test'], loc='upper left')
#plt.show()
# summarize history for loss plt.plot(history.history['loss']) plt.plot(history.history['val_loss']) plt.title('model loss')
ax2 = fig.add_subplot(1, 2, 2)
ax2.plot(history.history['loss'])
ax2.plot(history.history['val_loss'])
ax2.set_ylabel('loss')
ax2.set_xlabel('epoch')
ax2.legend(['train', 'test'], loc='upper left')
plt.show()
plt_loss(history)