# k_cross trainning
f = open(path.join(path.dirname(__file__),'..','record','temp.txt'), 'w')
# for n_epochs in [20,30,40,50,60,70,75,80,90,100,150,200,250,300]:
for n_epochs in [650,700,750,800]:
score_list_5chunk = []
confusion_matrix_5chunk = []
for train_chunk_number in range(5):
train_chunk_number = train_chunk_number + 1
train, test = k_cross(dataset, train_chunk_number)
X_train_text = train[:, 3: 103]
X_test_text = test[:, 3: 103]
X_train_senti = train[:, 103: 109]
X_test_senti = test[:, 103: 109]
X_train_lstm = train[:, 109: 237]
X_test_lstm = test[:, 109: 237]
X_train_dense = train[:, 237: 287]
X_test_dense = test[:, 237: 287]
X_train_decomdense = train[:, 287: 307]
X_test_decomdense = test[:, 287: 307]
Y_train = train[:, 0]
Y_test = test[:, 0]
for n_epochs in [15, 20, 30]:
score_list_5chunk = []
confusion_matrix_5chunk = []
for train_chunk_number in range(5):
train_chunk_number = train_chunk_number + 1
# train_text, test_text = k_cross(dataset, train_chunk_number)
X_train_lstm, X_test_lstm = k_cross_3(dynamic_lstm_dataset,
train_chunk_number)
# X_train_text = train_text[:, 3: 103]
# X_test_text = test_text[:, 3: 103]
# Y_train = train_text[:, label_index]
# Y_test = test_text[:, label_index]
train_text, test_text = k_cross(dataset, train_chunk_number)
Y_train, Y_test = k_cross(label, train_chunk_number)
X_train_cnn, X_test_cnn = k_cross_3(dynamic_lstm_dataset,
train_chunk_number)
X_train_cnn = X_train_cnn.reshape(
(X_train_cnn.shape[0], X_train_cnn.shape[1], X_train_cnn.shape[2],
1))
X_test_cnn = X_test_cnn.reshape(
(X_test_cnn.shape[0], X_test_cnn.shape[1], X_test_cnn.shape[2], 1))
X_train_text = train_text[:, 3:103]
X_test_text = test_text[:, 3:103]
# print(X_train_lstm.shape)
encoder = LabelEncoder()
encoder_label_train = encoder.fit_transform(Y_train)
print(dynamic_lstm_dataset.shape)
from keras.models import Model, Sequential
from keras.layers import Dense, Activation, Dropout, Embedding, Input
from keras.layers import LSTM
f = open(path.join(path.dirname(__file__), '..', 'record', 'temp.txt'), 'w')
for n_epochs in [20, 30, 40, 50, 60, 70, 75, 80, 90, 100, 150, 200, 250, 300]:
score_list_5chunk = []
confusion_matrix_5chunk = []
for train_chunk_number in range(5):
train_chunk_number = train_chunk_number + 1
train_text, test_text = k_cross(dataset, train_chunk_number)
X_train_lstm, X_test_lstm = k_cross_3(dynamic_lstm_dataset,
train_chunk_number)
X_train_text = train_text[:, 3:103]
X_test_text = test_text[:, 3:103]
Y_train = train_text[:, label_index]
Y_test = test_text[:, label_index]
# print(X_train_lstm.shape)
encoder = LabelEncoder()
encoder_label_train = encoder.fit_transform(Y_train)
dummy_Y_train = np_utils.to_categorical(encoder_label_train)
encoder_label_test = encoder.fit_transform(Y_test)
f = open(path.join(path.dirname(__file__), '..', 'record', 'temp.txt'), 'w')
# for n_epochs in [20,30,40,50,60,70,75,80,90,100,150,200,250,300]:
for n_epochs in [15, 20, 30]:
score_list_5chunk = []
confusion_matrix_5chunk = []
early_stopping = EarlyStopping(monitor='val_loss',
patience=12,
verbose=0,
mode='min')
for train_chunk_number in range(5):
train_chunk_number = train_chunk_number + 1
Y_train, Y_test = k_cross(label, train_chunk_number)
X_train, X_test = k_cross_3(user_sequence, train_chunk_number)
print(X_train.shape)
X_train = X_train.reshape(
(X_train.shape[0], X_train.shape[1], X_train.shape[2], 1))
X_test = X_test.reshape(
(X_test.shape[0], X_test.shape[1], X_test.shape[2], 1))
encoder = LabelEncoder()
encoder_label_train = encoder.fit_transform(Y_train)
dummy_Y_train = np_utils.to_categorical(encoder_label_train)
encoder_label_test = encoder.fit_transform(Y_test)
dummy_Y_test = np_utils.to_categorical(encoder_label_test)
# 二分类