def cnn_static(self):
x_train, y_train, x_test, y_test, vocabolary_dict = data_load_and_preproccess(
self.word_size, self.sequence_length)
embedding_weights = word_to_vector(np.vstack(
(x_train, x_test)), vocabolary_dict, self.embedding_dim,
self.min_word_count,
self.context_window_size)
x_train = np.stack([
np.stack([embedding_weights[word] for word in sentence])
for sentence in x_train
])
x_test = np.stack([
np.stack([embedding_weights[word] for word in sentence])
for sentence in x_test
])
#model building
model_input = Input(shape=(self.sequence_length, self.embedding_dim))
model = Dropout(self.drop_prob[0])(model_input)
multi_cnn_channel = []
for kernal in self.kernal_size:
conv_channel = Convolution1D(filters=self.filters,
kernel_size=kernal,
padding="valid",
activation="relu",
strides=1)(model)
conv_channel = MaxPooling1D(pool_size=2)(conv_channel)
conv_channel = Flatten()(conv_channel)
multi_cnn_channel.append(conv_channel)
model = Concatenate()(multi_cnn_channel)
model = Dropout(self.drop_prob[1])(model)
for dimension in self.hidden_dims:
model = Dense(dimension, activation="relu")(model)
model_output = Dense(1, activation="sigmoid")(model)
model = Model(model_input, model_output)
model.compile(loss="binary_crossentropy",
optimizer="adam",
metrics=["accuracy"])
print("Started Training : ")
model.fit(x_train,
y_train,
batch_size=self.batch_size,
epochs=self.epochs,
validation_data=(x_test, y_test),
verbose=2)
print("Training Completed")
self.model = model
def main(data_path, output_path):
X_trainS1, Y_train, X_valS1, Y_val = load_data(data_path)
epochs = 10
batch_size = 256
dropout_rate = 0.15
n_classes = 6
# 三个子模型的输入数据
main_input1 = Input(shape=(128, 3), name='main_input1')
def lstm_cell(main_input):
"""
基于DeepConvLSTM算法, 创建子模型
:param main_input: 输入数据
:return: 子模型
"""
sub_model = TimeDistributed(Dense(384),
input_shape=(128, 3))(main_input)
# sub_model = Flatten()(main_input)
print(sub_model)
sub_model = LSTM(256, return_sequences=True)(sub_model)
sub_model = LSTM(128, return_sequences=True)(sub_model)
sub_model = LSTM(128)(sub_model)
main_output = Dropout(dropout_rate)(sub_model)
return main_output
model = lstm_cell(main_input1)
model = Dropout(0.4)(model)
model = Dense(n_classes)(model)
model = BatchNormalization()(model)
output = Activation('softmax', name="softmax")(model)
model = Model([main_input1], output)
model.compile(loss='categorical_crossentropy',
optimizer='rmsprop',
metrics=['accuracy'])
print(model.summary())
# graph_path = os.path.join(output_path, "merged_model.png")
# plot_model(model, to_file=graph_path, show_shapes=True) # 绘制模型图
metrics = Metrics() # 度量FPR
history = model.fit(X_trainS1,
Y_train,
batch_size=batch_size,
validation_data=(X_valS1, Y_val),
epochs=epochs,
callbacks=[metrics]) # 增加FPR输出
model_path = os.path.join(output_path, "merged_dcl.h5")
model.save(model_path) # 存储模型
print(history.history)
def runResNet(epoch):
gloveD = 300
embMat = pickle.load(open('embMat' + str(gloveD) + '.pickle', 'rb'))
x_train_seq, y_train, x_val_seq, y_val = pickle.load(
open('trainValData.pickle', 'rb'))
act = 'relu'
inp = Input(shape=(300, ), dtype='int32')
e = Embedding(90461,
gloveD,
weights=[embMat],
input_length=300,
trainable=False)(inp)
#model = Flatten()(e)
model = Conv1D(filters=100,
kernel_size=2,
padding='valid',
activation=act,
strides=1)(e)
model = Dropout(0.5)(model)
model = GlobalMaxPooling1D()(model)
short = model
model = Dense(256, activation=act)(model)
model = Dropout(0.8)(model)
#short2 = model
#model = Add()([model, short])
#model = Activation(act)(model)
model = Dense(256, activation=act)(model)
model = Dropout(0.8)(model)
#short3 = model
#model = Add()([model, short2])
#model = Activation(act)(model)
model = Dense(256, activation=act)(model)
model = Dropout(0.8)(model)
#short4 = model
#model = Add()([model, short3])
#model = Activation(act)(model)
model = Dense(256, activation=act)(model)
model = Dropout(0.8)(model)
#short5 = model
#model = Add()([model,short4])
#model = Activation(act)(model)
model = Dense(100, activation=act)(model)
model = Dropout(0.8)(model)
model = Add()([model, short])
model = Activation(act)(model)
output = Dense(2, activation='softmax')(model)
model = Model(inputs=[inp], outputs=[output])
model.compile(loss="binary_crossentropy",
optimizer='adam',
metrics=['accuracy'])
filepath = "res_best_weights.{epoch:02d}-{val_acc:.4f}.hdf5"
checkpoint = ModelCheckpoint(filepath,
monitor='val_acc',
verbose=1,
save_best_only=True,
mode='max')
model.fit(x_train_seq,
y_train,
validation_data=(x_val_seq, y_val),
epochs=epoch,
batch_size=128,
verbose=1,
callbacks=[checkpoint])
def main(data_path, output_path):
X_trainS1, Y_train, X_valS1, Y_val = load_data(data_path)
epochs = 10
batch_size = 256
kernel_size = 3
pool_size = 2
dropout_rate = 0.15
n_classes = 6
f_act = 'relu'
# 三个子模型的输入数据
main_input1 = Input(shape=(128, 3), name='main_input1')
def cnn_lstm_cell(main_input):
"""
基于DeepConvLSTM算法, 创建子模型
:param main_input: 输入数据
:return: 子模型
"""
sub_model = Conv1D(512,
kernel_size,
input_shape=(128, 3),
activation=f_act,
padding='same')(main_input)
print('sub_model512:', sub_model)
sub_model = BatchNormalization()(sub_model)
sub_model = MaxPooling1D(pool_size=pool_size)(sub_model)
print('sub_model:', sub_model)
sub_model = Dropout(dropout_rate)(sub_model)
print('sub_model:', sub_model)
sub_model = Conv1D(64, kernel_size, activation=f_act,
padding='same')(sub_model)
print('sub_model64:', sub_model)
sub_model = BatchNormalization()(sub_model)
sub_model = MaxPooling1D(pool_size=pool_size)(sub_model)
print('sub_model:', sub_model)
sub_model = Dropout(dropout_rate)(sub_model)
print('sub_model:', sub_model)
sub_model = Conv1D(32, kernel_size, activation=f_act,
padding='same')(sub_model)
print('sub_model32:', sub_model)
sub_model = BatchNormalization()(sub_model)
sub_model = MaxPooling1D(pool_size=pool_size)(sub_model)
print('sub_model:', sub_model)
sub_model = LSTM(128, return_sequences=True)(sub_model)
print('sub_model128_1:', sub_model)
sub_model = LSTM(128, return_sequences=True)(sub_model)
print('sub_model128_2:', sub_model)
sub_model = LSTM(128)(sub_model)
print('sub_model128_3:', sub_model)
main_output = Dropout(dropout_rate)(sub_model)
print('main_output:', main_output)
return main_output
model = cnn_lstm_cell(main_input1)
model = Dropout(0.4)(model)
model = Dense(n_classes)(model)
model = BatchNormalization()(model)
output = Activation('softmax', name="softmax")(model)
model = Model([main_input1], output)
model.compile(loss='categorical_crossentropy',
optimizer='rmsprop',
metrics=['accuracy'])
# graph_path = os.path.join(output_path, "merged_model.png")
# plot_model(model, to_file=graph_path, show_shapes=True) # 绘制模型图
metrics = Metrics() # 度量FPR
history = model.fit(X_trainS1,
Y_train,
batch_size=batch_size,
validation_data=(X_valS1, Y_val),
epochs=epochs,
callbacks=[metrics]) # 增加FPR输出
model_path = os.path.join(output_path, "merged_dcl.h5")
model.save(model_path) # 存储模型
print(history.history)
tangential_strain_layer = (Dense(32, activation='relu'))(tangential_strain_layer)
tangential_strain_layer = Dropout(0.5)(tangential_strain_layer)
tangential_strain_layer = Flatten()(tangential_strain_layer)
merge = concatenate([energy_layer, maximum_amplitude_layer, radial_strain_layer, tangential_strain_layer])
model = (Dense(32, activation='relu'))(merge)
model = Dropout(0.5)(model)
model = (Dense(1, activation='sigmoid'))(model)
model = Model(inputs=[energy_input, maximum_amplitude_input, radial_strain_input, tangential_strain_input],
outputs=model)
model.compile(loss='binary_crossentropy', optimizer="rmsprop", metrics=['accuracy'])
print model.summary()
print "Training....................................................."
model.fit([energies, maximum_amplitudes, radial_strains, tangential_strains], train_labels, epochs=epochs, verbose=1,
batch_size=batch_size, validation_split=0.2)
print("--- train %s seconds ---" % (time.time() - start_time))
original_classified = model.predict([energies, maximum_amplitudes, radial_strains, tangential_strains],
batch_size=batch_size, verbose=0)
best_threshold = find_best_threshold(train_labels, original_classified)
if run_test:
print 'test'
for test in test_sets:
print test
test_reader = pd.read_csv(get_diff_path(test))
energy_test, maximum_amplitude_test, radial_strain_test, tangential_strain_test, test_labels = read_data(
test_reader)
opt = optimizers.Adam(
) # Default: lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=False,
model.compile(optimizer=opt,
loss=losses.binary_crossentropy,
metrics=['accuracy'])
print(get_available_gpus())
# Model Summary
model.summary()
# Start timer
start_time = time.time()
# Fit model
model.fit(x_train,
y_train,
batch_size=16,
validation_data=(x_valid, y_valid),
epochs=epochs,
shuffle=True,
verbose=2,
callbacks=[metrics])
# Get time
total_train_time = time.time() - start_time
print('Total training time in seconds: ')
print(total_train_time)
# Save model to file
model.save(os.path.join('.', 'trained_model.h5'))
# Plotting Epoch/accuracy
# print (metrics.acc)
# plt.plot(metrics.acc)
# plt.plot(metrics.val_acc,color='red')
# plt.xlabel('epochs')
def BiLSTM_model(filename,
train,
output,
X_train,
X_test,
word2ind,
maxWords,
y_train,
y_test,
ind2label,
validation=False,
X_valid=None,
y_valid=None,
word_embeddings=True,
pretrained_embedding="",
word_embedding_size=100,
maxChar=0,
char_embedding_type="",
char2ind="",
char_embedding_size=50,
lstm_hidden=32,
nbr_epochs=5,
batch_size=32,
dropout=0,
optimizer='rmsprop',
early_stopping_patience=-1,
folder_path="model_results",
gen_confusion_matrix=False):
"""
Build, train and test a BiLSTM Keras model. Works for multi-tasking learning.
The model architecture looks like:
- Words representations:
- Word embeddings
- Character-level representation [Optional]
- Dropout
- Bidirectional LSTM
- Dropout
- Softmax/CRF for predictions
:param filename: File to redirect the printing
:param train: Boolean if the model must be trained or not. If False, the model's wieght are expected to be stored in "folder_path/filename/filename.h5"
:param otput: "crf" or "softmax". Type of prediction layer to use
:param X_train: Data to train the model
:param X_test: Data to test the model
:param word2ind: Dictionary containing all words in the training data and a unique integer per word
:param maxWords: Maximum number of words in a sequence
:param y_train: Labels to train the model for the prediction task
:param y_test: Labels to test the model for the prediction task
:param ind2label: Dictionary where all labels for task 1 are mapped into a unique integer
:param validation: Boolean. If true, the validation score will be computed from 'X_valid' and 'y_valid'
:param X_valid: Optional. Validation dataset
:param y_valid: Optional. Validation dataset labels
:param word_embeddings: Boolean value. Add word embeddings into the model.
:param pretrained_embedding: Use the pretrained word embeddings.
Three values:
- "": Do not use pre-trained word embeddings (Default)
- False: Use the pre-trained embedding vectors as the weights in the Embedding layer
- True: Use the pre-trained embedding vectors as weight initialiers. The Embedding layer will still be trained.
:param word_embedding_size: Size of the pre-trained word embedding to use (100 or 300)
:param maxChar: The maximum numbers of characters in a word. If set to 0, the model will not use character-level representations of the words
:param char_embedding_type: Type of model to use in order to compute the character-level representation of words: Two values: "CNN" or "BILSTM"
:param char2ind: A dictionary where each character is maped into a unique integer
:param char_embedding_size: size of the character-level word representations
:param lstm_hidden: Dimentionality of the LSTM output space
:param nbr_epochs: Number of epochs to train the model
:param batch_size: Size of batches while training the model
:param dropout: Rate to apply for each Dropout layer in the model
:param optimizer: Optimizer to use while compiling the model
:param early_stopping_patience: Number of continuous tolerated epochs without improvement during training.
:param folder_path: Path to the directory storing all to-be-generated files
:param gen_confusion_matrix: Boolean value. Generated confusion matrices or not.
:return: The classification scores for both tasks.
"""
print("====== {0} start ======".format(filename))
end_string = "====== {0} end ======".format(filename)
# Create directory to store results
os.makedirs(folder_path + "/" + filename)
filepath = folder_path + "/" + filename + "/" + filename
# Set print outputs file
file, stdout_original = setPrintToFile("{0}.txt".format(filepath))
# Model params
nbr_words = len(word2ind) + 1
out_size = len(ind2label) + 1
best_results = ""
embeddings_list = []
inputs = []
# Input - Word Embeddings
if word_embeddings:
word_input = Input((maxWords, ))
inputs.append(word_input)
if pretrained_embedding == "":
word_embedding = Embedding(nbr_words,
word_embedding_size)(word_input)
else:
# Retrieve embeddings
embedding_matrix = word2VecEmbeddings(word2ind,
word_embedding_size)
word_embedding = Embedding(nbr_words,
word_embedding_size,
weights=[embedding_matrix],
trainable=pretrained_embedding,
mask_zero=False)(word_input)
embeddings_list.append(word_embedding)
# Input - Characters Embeddings
if maxChar != 0:
character_input = Input((
maxWords,
maxChar,
))
char_embedding = character_embedding_layer(char_embedding_type,
character_input, maxChar,
len(char2ind) + 1,
char_embedding_size)
embeddings_list.append(char_embedding)
inputs.append(character_input)
# Model - Inner Layers - BiLSTM with Dropout
embeddings = concatenate(embeddings_list) if len(
embeddings_list) == 2 else embeddings_list[0]
model = Dropout(dropout)(embeddings)
model = Bidirectional(
LSTM(lstm_hidden, return_sequences=True, dropout=dropout))(model)
model = Dropout(dropout)(model)
if output == "crf":
# Output - CRF
crfs = [[CRF(out_size), out_size]
for out_size in [len(x) + 1 for x in ind2label]]
outputs = [x[0](Dense(x[1])(model)) for x in crfs]
model_loss = [x[0].loss_function for x in crfs]
model_metrics = [x[0].viterbi_acc for x in crfs]
if output == "softmax":
outputs = [
Dense(out_size, activation='softmax')(model)
for out_size in [len(x) + 1 for x in ind2label]
]
model_loss = ['categorical_crossentropy' for x in outputs]
model_metrics = None
# Model
model = Model(inputs=inputs, outputs=outputs)
model.compile(loss=model_loss,
metrics=model_metrics,
optimizer=get_optimizer(optimizer))
print(model.summary(line_length=150), "\n\n\n\n")
# Training Callbacks:
callbacks = []
value_to_monitor = 'val_f1'
best_model_weights_path = "{0}.h5".format(filepath)
# 1) Classifition scores
classification_scores = Classification_Scores([X_train, y_train],
ind2label,
best_model_weights_path)
callbacks.append(classification_scores)
# 2) EarlyStopping
if early_stopping_patience != -1:
early_stopping = EarlyStopping(monitor=value_to_monitor,
patience=early_stopping_patience,
mode='max')
callbacks.append(early_stopping)
# Train
if train:
# Train the model. Keras's method argument 'validation_data' is referred as 'testing data' in this code.
hist = model.fit(X_train,
y_train,
validation_data=[X_test, y_test],
epochs=nbr_epochs,
batch_size=batch_size,
callbacks=callbacks,
verbose=2)
print()
print('-------------------------------------------')
print(
"Best F1 score:", early_stopping.best,
" (epoch number {0})".format(
1 + np.argmax(hist.history[value_to_monitor])))
# Save Training scores
save_model_training_scores("{0}".format(filepath), hist,
classification_scores)
# Print best testing classification report
best_epoch = np.argmax(hist.history[value_to_monitor])
print(classification_scores.test_report[best_epoch])
# Best epoch results
best_results = model_best_scores(classification_scores, best_epoch)
# Load weigths from best training epoch into model
save_load_utils.load_all_weights(model, best_model_weights_path)
# Create confusion matrices
if gen_confusion_matrix:
for i, y_target in enumerate(y_test):
# Compute predictions, flatten
predictions, target = compute_predictions(model, X_test, y_target,
ind2label[i])
# Generate confusion matrices
save_confusion_matrix(
target, predictions, list(ind2label[i].values()),
"{0}_task_{1}_confusion_matrix_test".format(
filepath, str(i + 1)))
# Validation dataset
if validation:
print()
print("Validation dataset")
print("======================")
# Compute classification report
for i, y_target in enumerate(y_valid):
# Compute predictions, flatten
predictions, target = compute_predictions(model,
X_valid,
y_target,
ind2label[i],
nbrTask=i)
# Only for multi-task
if len(y_train) > 1:
print("For task " + str(i + 1) + "\n")
print(
"===================================================================================="
)
print("")
print("With padding into account")
print(
metrics.flat_classification_report([target], [predictions],
digits=4))
print("")
print('----------------------------------------------')
print("")
print("Without the padding:")
print(
metrics.flat_classification_report([target], [predictions],
digits=4,
labels=list(
ind2label[i].values())))
# Generate confusion matrices
save_confusion_matrix(
target, predictions, list(ind2label[i].values()),
"{0}_task_{1}_confusion_matrix_validation".format(
filepath, str(i + 1)))
# Close file
closePrintToFile(file, stdout_original)
print(end_string)
return best_results
