Table 2.

Summary of deep learning models used in the studies reviewed with the time series-format input

Study	Proposed/adopted model(s)	Baselines	Evaluation metrics
Year 2019
[50]	LSTM	DNN	RMSE
Year 2020
[91]	Memory time-series network	AR, LRidge, LSVR, GP, VAR-MLP, GRU, LSTNet	RSE, R, RMSE, MAE
[32]	TCN	ARIMA, LSTM	MAPE
[11]	LSTM, GRU	RF, FFNN	MAE, RMSE, MAPE, RMSLE
[92]	WADC (CNN+LSTM+Attention)	ARIMA, SVR, Deep Regression, CNN, SAES, LSTM, GRU, LSTM-CNN, Deep&Cross Net	RMSE, MAE, MSLE
Year 2021
[22]	Multi-Output VP-RNN	HA, MA, LR, Poisson-RNN, VP-RNN	RMSE, MAE, $R^2$
[55]	LSTM	HW, kNN	MAE, RMSE
[51]	CQRNN	N/A	RMSE, MAE, $R^2$
[72]	FFNN	N/A	MSE, $R^2$
[15]	Bi-LSTM	RF, XGBoost, DNN, LSTM	MSE, RMSE, MAPE, $R^2$