ỨNG DỤNG MÔ HÌNH HỌC MÁY DỰ BÁO LƯU LƯỢNG ĐẾN HỒ TẢ TRẠCH

Abstract

In this study, the authors have proposed three regression models LR, RFR and LGBMR to predict the flow in Ta Trach reservoir of Thua Thien Hue province. The collected data set included the flow to the lake, the rainfall at 3 stations of Ta Trach, Thuong Nhat, and Nam Dong. The results showed that the LR model gave the best results with the criteria R2 of 0.97, RMSE of 58.31, MAE of 27.68. In addition, the RFR model also gave good results with a MAX-ERROR of 1587.43.