-
Notifications
You must be signed in to change notification settings - Fork 4
/
lstm.py
183 lines (152 loc) · 6.77 KB
/
lstm.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
from tensorflow.keras import Model
from sklearn.model_selection import StratifiedKFold
from tensorflow.keras import Model
from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint, ReduceLROnPlateau, CSVLogger
from tensorflow.keras.layers import *
from trys.custom_layer import ASPP
from load_data import *
from load_inv_data import *
from utils import *
from sklearn.metrics import *
train_lstm, y1, test_lstm, seq_len, _ = load_lstm_data()
train_lstm_inv, _, test_lstm_inv, _, _ = load_lstm_inv_data()
y = load_y()
sub = pd.read_csv('data/提交结果示例.csv')
def LSTM_FCN():
input = Input(shape=(train_lstm.shape[1], train_lstm.shape[2]), name="input_layer")
x = LSTM(64)(input)
x = Dropout(0.8)(x)
# y = Permute((2, 1))(input)
y = Conv1D(512, 8, padding='same', kernel_initializer='he_uniform')(input)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = Conv1D(256, 6, padding='same', kernel_initializer='he_uniform')(y)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = Conv1D(512, 4, padding='same', kernel_initializer='he_uniform')(y)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = GlobalAveragePooling1D()(y)
x = concatenate([x, y])
pred = Dense(19, activation='softmax')(x)
model = Model([input], pred)
return model
def LSTM_FCN_v2():
"""
线上：0.7605873015873017
:return:
"""
input = Input(shape=(train_lstm.shape[1], train_lstm.shape[2]), name="input_layer")
x = LSTM(64)(input)
x = Dropout(0.8)(x)
# y = Permute((2, 1))(input)
y = Conv1D(512, 8, padding='same', kernel_initializer='he_uniform')(input)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = Conv1D(256, 6, padding='same', kernel_initializer='he_uniform')(y)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = Conv1D(512, 4, padding='same', kernel_initializer='he_uniform')(y)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = ASPP(256, 3, activation=tf.nn.relu)(y)
y = GlobalAveragePooling1D()(y)
x = concatenate([x, y])
pred = Dense(19, activation='softmax')(x)
model = Model([input], pred)
return model
def base_boock(input):
x = LSTM(64)(input)
x = Dropout(0.8)(x)
# y = Permute((2, 1))(input)
y = Conv1D(512, 8, padding='same', kernel_initializer='he_uniform')(input)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = Conv1D(256, 6, padding='same', kernel_initializer='he_uniform')(y)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = Conv1D(512, 4, padding='same', kernel_initializer='he_uniform')(y)
y = BatchNormalization()(y)
y = Activation('relu')(y)
y = ASPP(256, 3, activation=tf.nn.relu)(y)
y = GlobalAveragePooling1D()(y)
x = concatenate([x, y])
return x
def LSTM_FCN_v3():
input_forward = Input(shape=(train_lstm.shape[1], train_lstm.shape[2]))
input_backward = Input(shape=(train_lstm.shape[1], train_lstm.shape[2]))
lstm_forward = base_boock(input_forward)
lstm_backward = base_boock(input_backward)
output = Concatenate(axis=-1)([lstm_forward, lstm_backward])
output = BatchNormalization()(Dropout(0.2)(Dense(256, activation='relu')(Flatten()(output))))
pred = Dense(19, activation='softmax')(output)
model = Model([input_forward, input_backward], pred)
return model
acc_scores = []
combo_scores = []
final_x = np.zeros((7292, 19))
proba_t = np.zeros((7500, 19))
kfold = StratifiedKFold(5, shuffle=True)
# 类别权重设置
class_weight = np.array([0.03304992, 0.09270433, 0.05608886, 0.04552935, 0.05965442,
0.04703785, 0.10175535, 0.03236423, 0.0449808, 0.0393582,
0.03236423, 0.06157433, 0.10065826, 0.03990675, 0.01727921,
0.06555129, 0.04731212, 0.03551838, 0.04731212])
for fold, (train_index, valid_index) in enumerate(kfold.split(train_lstm, y)):
print("{}train {}th fold{}".format('==' * 20, fold + 1, '==' * 20))
y_ = to_categorical(y, num_classes=19)
# model = LSTM_FCN()
model = LSTM_FCN_v3()
model.compile(loss='categorical_crossentropy',
optimizer='rmsprop',
metrics=['acc'])
model.summary()
plateau = ReduceLROnPlateau(monitor="val_acc",
verbose=1,
mode='max',
factor=0.5,
patience=15)
early_stopping = EarlyStopping(monitor='val_acc',
verbose=1,
mode='max',
patience=50)
checkpoint = ModelCheckpoint(f'models/fold{fold}.h5',
monitor='val_acc',
verbose=0,
mode='max',
save_best_only=True)
csv_logger = CSVLogger('../logs/log.csv', separator=',', append=True)
model.fit([train_lstm[train_index],
train_lstm_inv[train_index]],
y_[train_index],
epochs=500,
batch_size=256,
verbose=2,
shuffle=True,
class_weight=dict(enumerate((1 - class_weight) ** 3)),
validation_data=([train_lstm[valid_index],
train_lstm_inv[valid_index]],
y_[valid_index]),
callbacks=[plateau, early_stopping, checkpoint, csv_logger])
model.load_weights(f'models/fold{fold}.h5')
proba_x = model.predict([train_lstm[valid_index],
train_lstm_inv[valid_index]], verbose=0, batch_size=1024)
proba_t += model.predict([test_lstm, test_lstm_inv], verbose=0, batch_size=1024) / 5.
final_x[valid_index] += proba_x
oof_y = np.argmax(proba_x, axis=1)
score1 = accuracy_score(y[valid_index], oof_y)
# print('accuracy_score',score1)
score = sum(acc_combo(y_true, y_pred) for y_true, y_pred in zip(y[valid_index], oof_y)) / oof_y.shape[0]
print('accuracy_score', score1, 'acc_combo', score)
acc_scores.append(score1)
combo_scores.append(score)
print("acc_scores:", acc_scores)
print("combo_scores:", combo_scores)
print("5kflod mean acc score:{}".format(np.mean(acc_scores)))
print("5kflod mean combo score:{}".format(np.mean(combo_scores)))
sub.behavior_id = np.argmax(proba_t, axis=1)
sub.to_csv('result/lstm_acc{}_combo{}.csv'.format(np.mean(acc_scores), np.mean(combo_scores)), index=False)
pd.DataFrame(proba_t, columns=['pred_{}'.format(i) for i in range(19)]).to_csv(
'result/lstm_fcn_proba_t_{}.csv'.format(np.mean(acc_scores)), index=False)
pd.DataFrame(final_x, columns=['pred_{}'.format(i) for i in range(19)]).to_csv(
'result/lstm_fcn_proba_x_{}.csv'.format(np.mean(acc_scores)), index=False)