LSTM_ts: Long- Short Term Memory Model

In TSdeeplearning: Deep Learning Model for Time Series Forecasting

Long- Short Term Memory Model

Description

The LSTM function computes forecasted value with different forecasting evaluation criteria for long- short term memory model.

Usage

LSTM_ts(xt, xtlag = 4, uLSTM = 2, Drate = 0, nEpochs = 10,
Loss = "mse", AccMetrics = "mae",ActFn = "tanh",
Split = 0.8, Valid = 0.1)

Arguments

xt	Input univariate time series (ts) data.
xtlag	Lag of time series data.
uLSTM	Number of unit in LSTM layer.
Drate	Dropout rate.
nEpochs	Number of epochs.
Loss	Loss function.
AccMetrics	Metrics.
ActFn	Activation function.
Split	Index of the split point and separates the data into the training and testing datasets.
Valid	Validation set.

Details

Long short-term memory (LSTM) (Hochreiter and Schmidhuber, 1997) based RNN is designed to overcome the vanishing gradients problem while dealing with long term dependencies. In contrast to standard RNN, LSTM has this peculiar and unique inbuilt ability by maintaining a memory cell to determine which unimportant features should be forgotten and which important features should be remembered during the learning process (Jaiswal et al., 2022). An LSTM model analyses and captures both short-term and long-term temporal dependencies of a complex time series effectively due to its architecture of recurrent neural network and the memory function used in the hidden nodes.

Value

TrainFittedValue	Training Fitted value for given time series data.
TestPredictedValue	Final forecasted value of the LSTM model.
fcast_criteria	Different Forecasting evaluation criteria for LSTM model.

References

Cho, K., Van Merriënboer, B., Bahdanau, D. and Bengio, Y. (2014). On the properties of neural machine translation: Encoder-decoder approaches. arXiv preprint arXiv:1409.1259.

See Also

GRU, RNN

Examples

data("Data_Maize")
LSTM_ts(Data_Maize)

TSdeeplearning documentation built on Sept. 9, 2022, 9:05 a.m.