Khabat KhosraviNasrin Fathollahzadeh AttarSayed M. BateniChanghyun JunDongkyun KimMir Jafar Sadegh SafariSalim HeddamAitazaz Ahsan FarooqueSoroush AbolfathiSafari, Mir Jafar SadeghAbolfathi, SoroushJun, ChanghyunBateni, Sayed M.Kim, DongkyunAttar, NasrinKhosravi, Khabat2025-10-062024240584402405-844010.1016/j.heliyon.2024.e379652-s2.0-85206015921https://www.scopus.com/inward/record.uri?eid=2-s2.0-85206015921&doi=10.1016%2Fj.heliyon.2024.e37965&partnerID=40&md5=74f1fb2c37b31c7d3db3ee8be9e6afafhttps://gcris.yasar.edu.tr/handle/123456789/8150https://doi.org/10.1016/j.heliyon.2024.e37965Accurate prediction of daily river flow (Q<inf>t</inf>) remains a challenging yet essential task in hydrological modeling particularly crucial for flood mitigation and water resource management. This study introduces an advanced M5 Prime (M5P) predictive model designed to estimate Q<inf>t</inf> as well as one- and two-day-ahead river flow forecasts (i.e. Q<inf>t+1</inf> and Q<inf>t+2</inf>). The predictive performance of M5P ensembles incorporating Bootstrap Aggregation (BA) Disjoint Aggregating (DA) Additive Regression (AR) Vote (V) Iterative classifier optimizer (ICO) Random Subspace (RS) and Rotation Forest (ROF) were comprehensively evaluated. The proposed models were applied to a case study data in Tuolumne County US using a dataset comprising measured precipitation (P<inf>t</inf>) evaporation (E<inf>t</inf>) and Q<inf>t</inf>. A wide range of input scenarios were explored for predicting Q<inf>t</inf> Q<inf>t+1</inf> and Q<inf>t+2</inf>. Results indicate that P<inf>t</inf> and Q<inf>t</inf> significantly influence prediction accuracy. Notably relying solely on the most correlated variable (e.g. Q<inf>t-1</inf>) does not guarantee robust prediction of Q<inf>t</inf>. However extending the forecast horizon mitigates the influence of low-correlation input variables on model accuracy. Performance metrics indicate that the DA-M5P model achieves superior results with Nash-Sutcliff Efficiency of 0.916 and root mean square error of 23 m3/s followed by ROF-M5P BA-M5P AR-M5P AR-M5P RS-M5P V-M5P ICO-M5P and the standalone M5P model. The ensemble M5P modeling framework enhanced the predictive capability of the stand-alone M5P algorithm by 1.2 %–22.6 % underscoring its efficacy and potential for advancing hydrological forecasting. © 2024 Elsevier B.V. All rights reserved.Englishinfo:eu-repo/semantics/openAccessForecasting, Hybrid Machine Learning, M5p, Machine Learning, Predictive Model, River FlowM5PHybrid Machine LearningPredictive ModelRiver FlowMachine LearningForecastingArticle