Hai et al. (2025) A Novel Feature Engineering Method for Rainfall Prediction

Identification

• Journal: Lecture notes in networks and systems
• Year: 2025
• Date: 2025-10-27
• Authors: Nguyen Huu Hai, Tran Duc Quynh, Hoai Minh Le
• DOI: 10.1007/978-3-032-08384-5_11

Research Groups

• Vietnam National University of Agriculture, Hanoi, Vietnam
• Vietnam National University, Hanoi-International School, Hanoi, Vietnam
• Université de Lorraine, Metz, France

Short Summary

This paper introduces a novel feature engineering methodology designed to extract complex relationships from primary data for rainfall prediction. The proposed method significantly improves the performance of various machine learning models compared to using original features.

Objective

• To introduce and evaluate a new methodology for feature engineering from primary features to enhance the accuracy and reliability of rainfall prediction using machine learning models.

Study Configuration

• Spatial Scale: Not explicitly defined, focuses on a general methodology for rainfall prediction.
• Temporal Scale: Not explicitly defined, refers to general "rainfall forecast."

Methodology and Data

• Models used: A novel feature engineering methodology is proposed. Various machine learning models are used for evaluation, assessed via metrics such as Accuracy, F1-scores, and F1-score weighted.
• Data sources: Not explicitly detailed, but the method operates on "primary single features" and "original data" to derive more complex features.

Main Results

• The proposed feature engineering technique leads to a remarkable improvement in the performance of predictive models across various metrics (Accuracy, F1-scores, F1-score weighted) when compared to models trained using only original features.

Contributions

• Introduction of a novel feature engineering methodology specifically designed to better capture complex relationships and spatial-temporal dependencies within primary data for rainfall prediction.
• Demonstration of the significant performance enhancement achieved by this new feature engineering approach when applied to various machine learning models for rainfall forecasting.

Funding

• Not explicitly stated in the provided text.

Citation

@article{Hai2025Novel,
  author = {Hai, Nguyen Huu and Quynh, Tran Duc and Le, Hoai Minh},
  title = {A Novel Feature Engineering Method for Rainfall Prediction},
  journal = {Lecture notes in networks and systems},
  year = {2025},
  doi = {10.1007/978-3-032-08384-5_11},
  url = {https://doi.org/10.1007/978-3-032-08384-5_11}
}