-
Notifications
You must be signed in to change notification settings - Fork 0
/
models.py
151 lines (134 loc) · 6.18 KB
/
models.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
import preprocessing
import os
import tensorflow as tf
import numpy as np
import pandas as pd
import time
from tensorflow import keras
from tensorflow.keras import backend as K
from tensorflow.keras import layers
from tensorflow.keras.utils import normalize
from sklearn.preprocessing import OneHotEncoder ,LabelEncoder
from tensorflow.keras.callbacks import TensorBoard
from tensorflow.keras.backend import clear_session
import matplotlib.pyplot as plt
import logging
from sklearn.compose import ColumnTransformer
def set_tf_loglevel(level):
if level >= logging.FATAL:
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
if level >= logging.ERROR:
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
if level >= logging.WARNING:
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '1'
else:
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '0'
logging.getLogger('tensorflow').setLevel(level)
def weighted_categorical_crossentropy(weights):
"""
A weighted version of keras.objectives.categorical_crossentropy
Variables:
weights: numpy array of shape (C,) where C is the number of classes
Usage:
weights = np.array([0.5,2,10]) # Class one at 0.5, class 2 twice the normal weights, class 3 10x.
loss = weighted_categorical_crossentropy(weights)
model.compile(loss=loss,optimizer='adam')
"""
weights = K.variable(weights)
def loss(y_true, y_pred):
# scale predictions so that the class probas of each sample sum to 1
y_pred /= K.sum(y_pred, axis=-1, keepdims=True)
# clip to prevent NaN's and Inf's
y_pred = K.clip(y_pred, K.epsilon(), 1 - K.epsilon())
# calc
loss = y_true * K.log(y_pred) * weights
loss = -K.sum(loss, -1)
return loss
return loss
def load_data(csv_file, nr_classes):
print("Loading data ...")
data = np.loadtxt(csv_file, delimiter=',')
print("Data loaded...")
last_column = data[:, -1]
last_column = last_column.reshape(len(last_column), 1)
last_column = OneHotEncoder(sparse=False).fit_transform(last_column)
data[:, -1] = last_column[:, 0]
for i in range(nr_classes-1):
data = np.append(data, last_column[:, i+1].reshape(len(last_column[:, 1]), 1), axis=1)
X = data[:, :-nr_classes] # strip off indexes
y = data[:, -nr_classes:]
print(y)
#X = normalize(X, axis=0, order=1) # L2 norm?
return X, y
def visualize_data(csv_file, number_classes,number_samples):
X,y = load_data(csv_file, number_classes)
l = np.linspace(1000, number_classes*number_samples, 30, dtype=int)
print(l)
for index in l:
plt.figure()
plt.plot(X[index-1])
plt.title(str(y[index-1,0:number_classes]))
plt.show()
def CNN(training_data_path, list_classes, number_layers , window_size,
filter_number_conv = 32, dense_layer_size = None , epochs=20, pooling_size = 2 , validation_data_path = None, balanced = True):
clear_session()
set_tf_loglevel(logging.FATAL)
model_name = "{0}classes-CNN-{1}x{2}filt-{3}Dsize-{4}Poolsize-{5}W".format(list_classes,number_layers,
filter_number_conv, dense_layer_size, pooling_size, window_size)
while os.path.isdir(model_name):
try:
number = int(model_name.strip(".")[1])
number += 1
model_name = model_name.strip(".")[0] + "." +str(number)
except:
model_name = model_name + ".2"
tensorboard = TensorBoard(log_dir ='logs/{}'.format(model_name))
X_train,y_train = load_data(training_data_path, len(list_classes))
print(X_train.shape)
X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], 1))
print(X_train.shape)
if validation_data_path is not None:
X_validation, y_validation = load_data(validation_data_path, len(list_classes))
X_validation = np.reshape(X_validation, (X_validation.shape[0], X_validation.shape[1], 1))
model = tf.keras.Sequential()
for i in range(number_layers):
model.add(layers.Conv1D(filter_number_conv, 3, activation='relu'))
model.add(layers.MaxPooling1D(pool_size=2))
model.add(layers.Flatten())
if dense_layer_size is not None:
model.add(layers.Dense(dense_layer_size, activation = 'relu'))
model.add(layers.Dense(len(list_classes), activation='softmax'))
weights = []
for index in range(len(y_train[0])):
count = len([y for y in y_train[:, index] if y == 1])
weights.append(count/len(y_train[:, index]))
weights =np.array(weights)
print(weights)
model.compile(loss=weighted_categorical_crossentropy(weights), optimizer="adam",
metrics=['accuracy', 'Recall', 'Precision'])
if validation_data_path is not None:
model.fit(X_train, y_train, batch_size=32, epochs=epochs, shuffle=True, validation_data=(X_validation,y_validation),
callbacks=[tensorboard])
else:
model.fit(X_train, y_train, batch_size=32, epochs=epochs, shuffle=True, validation_split=0.1,
callbacks=[tensorboard])
model.summary()
model.save(model_name)
return model
def fullConnectedModel(number_classes):
model_name = "FullConnectedModel-3classes-12000samples-{}".format(str(time.time()))
X_train, y_train = load_data('training3-class-12000sam-2w-1.5ov.csv', number_classes)
print(X_train)
print(y_train)
tensorboard = TensorBoard(log_dir='logs/{}'.format(model_name))
model = tf.keras.Sequential()
model.add(layers.Dense(50, input_dim = X_train.shape[1], activation = 'relu'))
model.add(layers.Dense(3, activation= 'softmax'))
model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy','Recall','Precision'])
model.fit(X_train, y_train, batch_size=32, epochs=100, validation_data=0.1,
callbacks=[tensorboard], shuffle=True)
model.save(model_name)
return model
base_path = "/Volumes/KESU/Datasets/TUH_EEG_Seizure_Corpus/v1.5.0/edf/train/FP2/csv_files/"
CNN(base_path + "['absz', 'bckg']-20000samp-2w-250-fs-TrainingLimited.csv",
["absz","bckg"], 4, 2, filter_number_conv=64, pooling_size=3, epochs=10, validation_data_path= base_path + "['absz', 'bckg']-20000samp-2w-250-fs-ValidationLimited.csv", balanced=False)