keras_rnn: Train Recurrent Neural Network with Keras

In thfuchs/tsRNN: Time Series Forecasting using Recurrent Neural Networks by Keras

Description

Currently supports "simple unit", "gated recurrent unit" (GRU) and "Long-Short Term Memory" (LSTM) using Keras framework (with TensorFlow backend)

Usage

keras_rnn(
  X,
  Y,
  model_type,
  tsteps,
  n_epochs = 200,
  n_units = 32,
  loss = "mse",
  metrics = NULL,
  dropout_in_test = FALSE,
  optimizer = optimizer_rmsprop(),
  dropout = 0,
  recurrent_dropout = 0,
  history = FALSE,
  live_plot = FALSE
)

Arguments

X	list of "train", "val", and "test" with 3D (keras) arrays
Y	list of "train", "val", and "test" with 2D (keras) arrays
model_type	One of "simple", "gru" and "lstm"
tsteps	number of time steps for keras input shape
n_epochs	default 200
n_units	32 (currently fixed)
loss	default "mse"
metrics	default NULL
dropout_in_test	apply dropout during training only (default) or during testing also? Required for dropout-based prediction intervals (bayesian RNN)
optimizer	from keras, e.g. optimizer_rmsprop
dropout	dropout rate
recurrent_dropout	Dropout rate applied to reccurent layer. Default 0
history	in addition to model, return model history? Beware that output changes from model to list(model, history) if history = TRUE
live_plot	plot loss and validation metric during training? False by default

Value

Keras model by default (history = FALSE) else list with Keras model and history

References

• Chollet, Francois and others (2015). Keras. https://keras.io

• Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9 (8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735

• Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014). Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling. https://arxiv.org/pdf/1412.3555

See Also

Keras Documentation

Examples

data <- tsRNN::DT_apple
data[, value_lag1 := data.table::shift(value, type = "lag", n = 1)]
data <- data[!is.na(get(paste0("value_lag1")))]

nn_arrays <- ts_nn_preparation(data, tsteps = 1L, length_val = 6L, length_test = 6L)
keras_rnn(nn_arrays$x, nn_arrays$y, model_type = "simple", tsteps = 1, n_epochs = 20)

# return model and history
result <- keras_rnn(
  nn_arrays$x, nn_arrays$y, model_type = "simple", tsteps = 1, n_epochs = 20, history = TRUE
)

result$model
result$history

## Not run: 
# Plot result
plot(result$history)

## End(Not run)

thfuchs/tsRNN documentation built on April 17, 2021, 11:03 p.m.