Determining meteorological factors influencing dengue cases in the National Capital Region using machine learning algorithms

Authors

Rica Mae Lumiguid Department of Physical Sciences, Polytechnic University of the Philippines
Rosenel Dahipon National Institute of Physics, University of the Philippines Diliman and Department of Physical Sciences, Polytechnic University of the Philippines
Rhenish Simon Physics Department, De La Salle University, Manila

Abstract

We use Principal Component Regression (PCR) and Random Forest Regression (RFR) models to identify the meteorological factors most influential on weekly DENV cases from 2009 to 2019 in the National Capital Region. The dataset split into a training set and a test set with an optimal ratio of 70% and 30%, respectively. The training set is utilize to train both the PCR and RFR models. In both the PCR and RFR models, maximum temperature, rainfall, minimum temperature, and relative humidity were identified as the top four important features influencing DENV cases, thereby enhancing mosquito reproduction and survival rates, and consequently increasing DENV cases. Additionally, wind direction ranked as the least important in both models. Moreover, the RFR outperformed the PCR, explaining 60.4% of the variability in DENV cases using meteorological factors.

Downloads

Issue

2024: Proceedings of the 42nd Samahang Pisika ng Pilipinas Physics Conference

Article ID

SPP-2024-1E-02

Section

Complex Systems and Data Analytics

Published

2024-06-26

How to Cite

[1]

RM Lumiguid, R Dahipon, and R Simon, Determining meteorological factors influencing dengue cases in the National Capital Region using machine learning algorithms, Proceedings of the Samahang Pisika ng Pilipinas 42, SPP-2024-1E-02 (2024). URL: https://proceedings.spp-online.org/article/view/SPP-2024-1E-02.