Three Steps towards Better Forecasting for Streamflow Deep Learning

Elevating the accuracy of streamflow forecasting has always been a challenge. This paper proposes three-step artificial intelligence model improvement for forecasting. Step 1 uses long short-term memory (LSTM), an on conventional neural network (ANN). 2 performs multi-step ahead while establishing rates change as new approach. 3 further improves through three different kinds optimization algorithms. The Stormwater and Road Tunnel project in Kuala Lumpur is study area. Historical rainfall data 14 years at 11 telemetry stations are obtained to forecast flow confluence located next control center. reveals that LSTM better than ANN with R 0.9055, MSE 17,8532, MAE 1.4365, NSE 0.8190 RMSE 5.3695. unveils outperforms rest = 0.9545, 8.9746, 0.5434, 0.9090 2.9958. Finally, Stage 0.9757, 4.7187, 0.4672, 0.9514 2.1723 bat-LSTM hybrid algorithm. shows δQ consistently yielded promising results metaheuristic algorithms able yield additional model’s results.