GRU_ts {TSdeeplearning}

Gated Recurrent Unit Model

Description

The GRU function computes forecasted value with different forecasting evaluation criteria for gated recurrent unit model.

Usage

GRU_ts(xt, xtlag = 4, uGRU = 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.

uGRU

Number of unit in GRU 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

The gated recurrent unit (GRU) was introduced by Cho et al.(2014). A GRU is part of a specific model of recurrent neural network that intends to use connections through a sequence of nodes to perform machine learning tasks associated with memory and clustering. Its internal structure is simpler and, therefore, it is also easier to train, as less calculation is required to upgrade the internal states. The update port controls the extent to which the state information from the previous moment is retained in the current state, while the reset port determines whether the current state should be combined with the previous information. Gated recurrent units help to adjust neural network input weights to solve the vanishing gradient problem that is a common issue with recurrent neural networks.

Value

TrainFittedValue

Training Fitted value for given time series data.

TestPredictedValue

Final forecasted value of the GRU model.

fcast_criteria

Different Forecasting evaluation criteria for GRU 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

LSTM, RNN

Examples

data("Data_Maize")
GRU_ts(Data_Maize)