def __init__(self, UCR_dataset_name ):
super(SingleLayerConfig, self).__init__()
X_train, y_train, X_test, y_test, series_size = get_dataset_with_series_size(dataset=UCR_dataset_name)
self.X_train = X_train.reshape((-1, series_size, 1))
self.X_test = X_test.reshape((-1, series_size, 1))
y_train = y_train.reshape((-1, 1))
y_test = y_test.reshape((-1, 1))
self.y_train = one_hot_to_int(y_train)
self.y_test = one_hot_to_int(y_test)
# Input data
self.train_count = len(self.X_train) # 7352 training series
self.test_data_count = len(self.X_test) # 2947 testing series
self.n_steps = len(self.X_train[0]) # 128 time_steps per series
# Trainging
self.learning_rate = 0.005
self.lambda_loss_amount = 0.0015
self.training_epochs = 100
self.batch_size = 1500
# LSTM structure
self.n_inputs = len(self.X_train[0][0])
self.n_hidden = 28 # nb of neurons inside the neural network
self.n_classes = len(self.y_train[0]) # Final output classes
self.model_name = "single_lstm_" + UCR_dataset_name
self.log_folder_suffix = self.attach_log_suffix()
self.logs_path = "/tmp/LSTM_logs/"+self.log_folder_suffix
self.tensor_board_logging_enabled = True
self.tensorboard_cmd = "tensorboard --logdir="+ self.logs_path
self.model_desc_attched_string = self.attach_mdoel_desc()
self.matplot_lib_enabled = True
self.matplot_lib_for_accuracy =True
self.matplot_lib_for_single_ybundle=False
def one_hot_to_int(y):
"""convert label from dense to one hot
argument:
label: ndarray dense label ,shape: [sample_num,1]
return:
one_hot_label: ndarray one hot, shape: [sample_num,n_class]
"""
# e.g.: [[5], [0], [3]] --> [[0, 0, 0, 0, 0, 1], [1, 0, 0, 0, 0, 0], [0, 0, 0, 1, 0, 0]]
y = y.reshape(len(y))
n_values = int(np.max(y)) + 1
return np.eye(n_values)[np.array(y, dtype=np.int32)] # Returns FLOATS
################################## load data and config ##################################
X_train, y_train, X_test, y_test, series_size = get_dataset_with_series_size(
dataset='ElectricDevices')
X_train = X_train.reshape((-1, series_size, 1))
X_test = X_test.reshape((-1, series_size, 1))
y_train = y_train.reshape((-1, 1))
y_test = y_test.reshape((-1, 1))
y_train = one_hot_to_int(y_train)
y_test = one_hot_to_int(y_test)
class SingleLayerConfig(Config):
"""
define a class to store parameters,
the input should be feature mat of training and testing
"""
def __init__(self):
super(SingleLayerConfig, self).__init__()
# New shape: a list of lenght "time_step" containing tensors of shape [batch_size, n_hidden]
hidden = get_stacked_highway_LSTM_layers(hidden, config.n_inputs,
config.n_hidden, config,
keep_prob_for_dropout)
# Final fully-connected activation logits
# Get the last output tensor of the inner loop output series, of shape [batch_size, n_classes]
last_hidden = tf.nn.dropout(hidden[-1], keep_prob_for_dropout)
last_logits = relu_fc([last_hidden], config.n_hidden, config.n_classes,
config)[0]
return last_logits
################################## load data and config ##################################
X_train, y_train, X_test, y_test, series_size = get_dataset_with_series_size(
dataset='synthetic_control')
#X_train, y_train, X_test, y_test, series_size = get_dataset_with_series_size(dataset='ElectricDevices')
X_train = X_train.reshape((-1, series_size, 1))
X_test = X_test.reshape((-1, series_size, 1))
y_train = y_train.reshape((-1, 1))
y_test = y_test.reshape((-1, 1))
y_train = one_hot_to_int(y_train)
y_test = one_hot_to_int(y_test)
class HighwayLSTMOnUCRConfig(Config):
def __init__(self):
super(HighwayLSTMOnUCRConfig, self).__init__()
self.train_count = len(X_train) # 7352 training series
self.test_data_count = len(X_test) # 2947 testing series