Predicting Energy Consumption Trends in the Philippines (2025-2030) Using Machine Learning Models

Document Types

Paper Presentation

School Name

De La Salle University

Track or Strand

Science, Technology, Engineering, and Mathematics (STEM)

Research Advisor (Last Name, First Name, Middle Initial)

Gue, Ivan Henderson, V.

Start Date

23-6-2025 3:30 PM

End Date

23-6-2025 5:00 PM

Zoom Link/ Room Assignment

EKR 405

Abstract/Executive Summary

Energy security remains a significant concern for the Philippines amid rising demand, growing population, and continuous modernization. With the history of the Philippines’ problems in energy, attention to energy is necessary. With this, the study aims to forecast the Philippines' possible annual energy consumption from 2025 to 2030, considering population growth as a potential variable influencing the Philippines’ projected energy consumption trends. The study employs a machine learning approach using the linear regression model, which was the model with the best performance in forecasting future energy consumption based on MATLAB’s Linear Regression Learner app. Prior to forecasting, historical data from 2010 to 2023 were analyzed in the study's correlation analysis to assess the relationship between energy consumption and population growth, revealing a strong positive linear relationship (r = 0.9). The linear regression model, on the other hand, showed high predictive accuracy ( R² = 0.93, RMSE = 4060.9 GWh, MAPE = 3.9%) with results projecting an increase in energy consumption from 109,750 GWh in 2024 to 122,710 GWh by 2030. The comparative analysis between DOE’s projected energy supply data and the study’s forecasted energy consumption indicates supply exceeding demand by 43,286 GWh by 2030. These findings suggest that DOE’s estimated energy supply is sufficient to meet the country’s needs through 2030.

Keywords

Philippine energy consumption; Philippine energy supply; energy consumption forecasting; machine learning forecasting; machine learning

Research Theme (for Paper Presentation and Poster Presentation submissions only)

Sustainability, Environment, and Energy (SEE)

Statement of Originality

yes

This document is currently not available here.

Share

COinS
 
Jun 23rd, 3:30 PM Jun 23rd, 5:00 PM

Predicting Energy Consumption Trends in the Philippines (2025-2030) Using Machine Learning Models

Energy security remains a significant concern for the Philippines amid rising demand, growing population, and continuous modernization. With the history of the Philippines’ problems in energy, attention to energy is necessary. With this, the study aims to forecast the Philippines' possible annual energy consumption from 2025 to 2030, considering population growth as a potential variable influencing the Philippines’ projected energy consumption trends. The study employs a machine learning approach using the linear regression model, which was the model with the best performance in forecasting future energy consumption based on MATLAB’s Linear Regression Learner app. Prior to forecasting, historical data from 2010 to 2023 were analyzed in the study's correlation analysis to assess the relationship between energy consumption and population growth, revealing a strong positive linear relationship (r = 0.9). The linear regression model, on the other hand, showed high predictive accuracy ( R² = 0.93, RMSE = 4060.9 GWh, MAPE = 3.9%) with results projecting an increase in energy consumption from 109,750 GWh in 2024 to 122,710 GWh by 2030. The comparative analysis between DOE’s projected energy supply data and the study’s forecasted energy consumption indicates supply exceeding demand by 43,286 GWh by 2030. These findings suggest that DOE’s estimated energy supply is sufficient to meet the country’s needs through 2030.

https://animorepository.dlsu.edu.ph/conf_shsrescon/2025/paper_see/13