Diana, Cyntia Riska (2025) Model Prediksi Curah Hujan Menggunakan Machine Learning dengan Memanfaatkan Data Remote. Other thesis, Universitas Islam Riau.

Text skripsi_203510038_watermark.pdf - Submitted Version Restricted to Registered users only Download (2MB) | Request a copy

Abstract

This research aims to develop a rainfall prediction model by utilising Machine Learning algorithms, specifically Random Forest and XGBoost algorithm as a comparison, by utilising remote sensing data and meteorological data from the Meteorology Climatology and Geophysics Agency (BMKG) in Pekanbaru City, Riau for the period January 2010 to December 2023. This research is motivated by the negative impacts of high rainfall, such as infrastructure damage and transport disruption, as well as challenges in traditional rainfall forecasting. With technological advances such as IoT and remote sensing, this research aims to improve the accuracy of rainfall prediction through Machine Learning models that can identify complex patterns in meteorological data. The methodology used includes data collection, pre-processing, model building, model evaluation, and hyperparameter tuning. The results showed that after hyperparameter tuning, the Random Forest model of scenario 70 : 30 became the most optimal with the highest R² value and the lowest MAE and MSE compared to XGBoost, namely with an R² value of 0.9066, MAE 0.8364 and MSE 2.0631. It is expected that the results of this study can make an important contribution to early warning systems and risk mitigation associated with rainfall in urban areas.

Actions (login required)

View Item