forked from keras-team/keras
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtest_loss_weighting.py
174 lines (138 loc) · 6.39 KB
/
test_loss_weighting.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
from __future__ import absolute_import
from __future__ import print_function
import pytest
import numpy as np
from keras import backend as K
from keras.utils.test_utils import get_test_data
from keras.models import Sequential, Model
from keras.layers import Dense, Activation, GRU, TimeDistributed, Input
from keras.utils import np_utils
from numpy.testing import assert_almost_equal, assert_array_almost_equal
num_classes = 10
batch_size = 128
epochs = 15
weighted_class = 5
high_weight = 10
train_samples = 5000
test_samples = 1000
timesteps = 3
input_dim = 10
loss = 'mse'
standard_weight = 1
standard_score_sequential = 0.5
decimal_precision = {
'cntk': 2,
'theano': 6,
'tensorflow': 6
}
def _get_test_data():
np.random.seed(1337)
(x_train, y_train), (x_test, y_test) = get_test_data(num_train=train_samples,
num_test=test_samples,
input_shape=(input_dim,),
classification=True,
num_classes=num_classes)
int_y_test = y_test.copy()
int_y_train = y_train.copy()
# convert class vectors to binary class matrices
y_train = np_utils.to_categorical(y_train, num_classes)
y_test = np_utils.to_categorical(y_test, num_classes)
test_ids = np.where(int_y_test == np.array(weighted_class))[0]
class_weight = dict([(i, standard_weight) for i in range(num_classes)])
class_weight[weighted_class] = high_weight
sample_weight = np.ones((y_train.shape[0])) * standard_weight
sample_weight[int_y_train == weighted_class] = high_weight
return ((x_train, y_train), (x_test, y_test),
(sample_weight, class_weight, test_ids))
def create_sequential_model():
model = Sequential()
model.add(Dense(32, input_shape=(input_dim,)))
model.add(Activation('relu'))
model.add(Dense(num_classes))
model.add(Activation('softmax'))
return model
def create_temporal_sequential_model():
model = Sequential()
model.add(GRU(32, input_shape=(timesteps, input_dim), return_sequences=True))
model.add(TimeDistributed(Dense(num_classes)))
model.add(Activation('softmax'))
return model
def test_sequential_class_weights():
model = create_sequential_model()
model.compile(loss=loss, optimizer='rmsprop')
((x_train, y_train), (x_test, y_test),
(sample_weight, class_weight, test_ids)) = _get_test_data()
model.fit(x_train, y_train, batch_size=batch_size,
epochs=epochs // 3, verbose=0,
class_weight=class_weight,
validation_data=(x_train, y_train, sample_weight))
model.fit(x_train, y_train, batch_size=batch_size,
epochs=epochs // 2, verbose=0,
class_weight=class_weight)
model.fit(x_train, y_train, batch_size=batch_size,
epochs=epochs // 2, verbose=0,
class_weight=class_weight,
validation_split=0.1)
model.train_on_batch(x_train[:32], y_train[:32],
class_weight=class_weight)
score = model.evaluate(x_test[test_ids, :], y_test[test_ids, :], verbose=0)
assert(score < standard_score_sequential)
def test_sequential_sample_weights():
model = create_sequential_model()
model.compile(loss=loss, optimizer='rmsprop')
((x_train, y_train), (x_test, y_test),
(sample_weight, class_weight, test_ids)) = _get_test_data()
model.fit(x_train, y_train, batch_size=batch_size,
epochs=epochs // 3, verbose=0,
sample_weight=sample_weight)
model.fit(x_train, y_train, batch_size=batch_size,
epochs=epochs // 3, verbose=0,
sample_weight=sample_weight,
validation_split=0.1)
model.train_on_batch(x_train[:32], y_train[:32],
sample_weight=sample_weight[:32])
model.test_on_batch(x_train[:32], y_train[:32],
sample_weight=sample_weight[:32])
score = model.evaluate(x_test[test_ids, :], y_test[test_ids, :], verbose=0)
assert(score < standard_score_sequential)
def test_sequential_temporal_sample_weights():
((x_train, y_train), (x_test, y_test),
(sample_weight, class_weight, test_ids)) = _get_test_data()
temporal_x_train = np.reshape(x_train, (len(x_train), 1, x_train.shape[1]))
temporal_x_train = np.repeat(temporal_x_train, timesteps, axis=1)
temporal_x_test = np.reshape(x_test, (len(x_test), 1, x_test.shape[1]))
temporal_x_test = np.repeat(temporal_x_test, timesteps, axis=1)
temporal_y_train = np.reshape(y_train, (len(y_train), 1, y_train.shape[1]))
temporal_y_train = np.repeat(temporal_y_train, timesteps, axis=1)
temporal_y_test = np.reshape(y_test, (len(y_test), 1, y_test.shape[1]))
temporal_y_test = np.repeat(temporal_y_test, timesteps, axis=1)
temporal_sample_weight = np.reshape(sample_weight, (len(sample_weight), 1))
temporal_sample_weight = np.repeat(temporal_sample_weight, timesteps, axis=1)
model = create_temporal_sequential_model()
model.compile(loss=loss, optimizer='rmsprop',
sample_weight_mode='temporal')
model.fit(temporal_x_train, temporal_y_train, batch_size=batch_size,
epochs=epochs // 3, verbose=0,
sample_weight=temporal_sample_weight)
model.fit(temporal_x_train, temporal_y_train, batch_size=batch_size,
epochs=epochs // 3, verbose=0,
sample_weight=temporal_sample_weight,
validation_split=0.1)
model.train_on_batch(temporal_x_train[:32], temporal_y_train[:32],
sample_weight=temporal_sample_weight[:32])
model.test_on_batch(temporal_x_train[:32], temporal_y_train[:32],
sample_weight=temporal_sample_weight[:32])
score = model.evaluate(temporal_x_test[test_ids], temporal_y_test[test_ids],
verbose=0)
assert(score < standard_score_sequential)
def test_class_weight_wrong_classes():
model = create_sequential_model()
model.compile(loss=loss, optimizer='rmsprop')
((x_train, y_train), (x_test, y_test),
(sample_weight, class_weight, test_ids)) = _get_test_data()
del class_weight[1]
with pytest.raises(ValueError):
model.fit(x_train, y_train,
epochs=0, verbose=0, class_weight=class_weight)
if __name__ == '__main__':
pytest.main([__file__])