We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Nonlinear Model for Precipitation Forecasting in Northern Iraq using Machine Learning Algorithms.
- Authors
Al-Hashimi, Muzahem M.; Hayawi, Heyam A. A.
- Abstract
Precipitation forecasting is a challenging task in meteorology, especially in Iraq, where the complex and nonlinear nature of precipitation requires powerful computational capabilities. Our study focuses on monthly rainfall in five northern provinces of Iraq: Nineveh, Dahuk, Erbil, Kirkuk, and Al-Sulaymaniyah. We propose bagging ensemble learning model incorporating Random Trees (RT), Locally Weighted Learning (LWL), and k-Nearest Neighbor (kNN) as base models. We utilize meteorological data and evaluate various input parameters. The results, measured using MSE, MAE, and RMSE, demonstrate the effectiveness and efficiency of our approach. The BG-LWL-RT model performs best in precipitation forecasting for Dahuk, Erbil, and Kirkuk, while the BG-LWL-kNN model achieves the highest accuracy for predicting precipitation in the Nineveh and Al-Sulaymaniyah datasets.
- Subjects
IRAQ; DIHOK (Iraq); MACHINE learning; PRECIPITATION forecasting; K-nearest neighbor classification
- Publication
International Journal of Mathematics & Computer Science, 2024, Vol 19, Issue 1, p171
- ISSN
1814-0424
- Publication type
Article