Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Rainfall Prediction Using Hybrid Model DWT-NARX-LSTM Based on Correlative Feature Selection

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
33c39f61-bba1-4617-a3b0-17d045b9d28d
Published
2026-10-30

Universitas Dian Nuswantoro conducted a study on rainfall prediction using a hybrid deep learning model that integrates Discrete Wavelet Transform (DWT), a Nonlinear Autoregressive with Exogenous Input (NARX) structure, and Long Short-Term Memory (LSTM). Giyarto, Muljono, and Santoso, Heru Agus, participated as researchers in this activity. The study focused on rainfall prediction because rainfall is an important variable in determining seasons and managing water resources. The source also states that prediction of weather elements, including rainfall, temperature, wind speed, and solar radiation, is beneficial for human life.

The study addressed the importance of rainfall prediction accuracy for hydrometeorological disaster mitigation and decision-making in the agricultural and infrastructure sectors. The hybrid model was enriched through correlative feature selection and feature engineering. The reported feature engineering techniques included lag, rolling mean, and rolling standard deviation, which were used to capture dynamic patterns in time series. Model evaluation was carried out using RMSE, MAE, and the Pearson correlation coefficient.

The reported evaluation results were an RMSE value of 3.4913, an MAE value of 2.5809, and a correlation of 0.9443. The study states that these results indicate that the model successfully provides rainfall predictions with high accuracy. The source describes the study as contributing to the development of an adaptive and precise rainfall prediction system. This institutional research activity presents the modelling components, feature techniques, evaluation metrics, and findings reported in the source. Further information is needed about the study location, rainfall data source, study period, and implementation details.

Reference https://www.scopus.com/pages/publications/105031589839?origin=resultslist