EUVIMEDEUVIMED entdecken
Crossref

Rainfall prediction in Maharashtra using machine learning models: a comparative evaluation of linear, ridge and support vector regression approaches

Rajesh Jadhav, Dhananjay Kharat, Sudarshan Talegaonkar, Pooja Pawar, Uday Patil

Disaster Advances · 30.06.2026 · 42

DOI: 10.25303/198da42054 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Disaster Advances, online erschienen am 30.06.2026. Verfasst von Rajesh Jadhav und 4 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate rainfall prediction is essential for effective flood forecasting, water resource planning and disaster mitigation, particularly in monsoon-dependent regions such as Maharashtra, India. This study explores the application of machine learning models Linear Regression, Ridge Regression and Support Vector Regression (SVR) for rainfall prediction using historical meteorological data obtained from the IndiaWRIS portal. The dataset underwent rigorous preprocessing including normalization and feature creation using lagged rainfall attributes to capture temporal dependencies. Model performance was evaluated using key statistical metrics such as R², RMSE, MAE and NSE, complemented by graphical analyses including observed versus predicted plots, error distributions and scatter density visualization. Results demonstrated that while all models achieved high predictive accuracy (R² > 0.98), the SVR model significantly outperformed the linear models with an R² of 0.995 and correlation coefficient (r) of 0.998 during the testing phase. The SVR’s ability to handle nonlinear rainfall relationships led to minimal residual errors and improved generalization across all phases. These findings highlight SVR’s robustness and its suitability for real-time rainfall forecasting applications. The proposed approach can serve as a foundation for developing intelligent flood warning systems and enhancing hydrological decision-support frameworks for the Maharashtra region.

Zitieren

APA

Jadhav, R., Kharat, D., Talegaonkar, S., Pawar, P., & Patil, U. (2026). Rainfall prediction in Maharashtra using machine learning models: a comparative evaluation of linear, ridge and support vector regression approaches. Disaster Advances, 42. https://doi.org/10.25303/198da42054

Vancouver

Jadhav R, Kharat D, Talegaonkar S, Pawar P, Patil U. Rainfall prediction in Maharashtra using machine learning models: a comparative evaluation of linear, ridge and support vector regression approaches. Disaster Advances. 2026:42. doi:10.25303/198da42054

BibTeX
@article{Jadhav2026,
  author = {Jadhav, Rajesh and Kharat, Dhananjay and Talegaonkar, Sudarshan and Pawar, Pooja and Patil, Uday},
  title = {Rainfall prediction in Maharashtra using machine learning models: a comparative evaluation of linear, ridge and support vector regression approaches},
  journal = {Disaster Advances},
  year = {2026},
  pages = {42},
  doi = {10.25303/198da42054}
}
Kostenlos starten →