Improvement of SVR-Based Drought Forecasting Models using Wavelet Pre-Processing Technique

摘要: Drought is a damaging natural hazard due to the lack of precipitation from expected amount for period time. Mitigations are required reduced its impact. Due difficulty in determining onset and offset droughts, accurate drought forecasting approaches risk management. Given growing use machine learning field, Wavelet-Boosting Support Vector Regression (W-BS-SVR) was proposed at Langat River Basin, Malaysia. Monthly rainfall, mean temperature evapotranspiration years 1976 - 2015 were used compute Standardized Precipitation Evapotranspiration Index (SPEI) this study, producing SPEI-1, SPEI-3 SPEI-6. The 1-month lead time SPEIs capability W-BS-SVR model compared with (SVR) Boosting-Support (BS-SVR) models using Root Mean Square Error (RMSE), Absolute (MAE), coefficient determination (R 2 ) Adjusted R . results demonstrated that provides higher accuracy prediction Basin.