This document discusses using machine learning techniques to analyze factors that influence particulate matter (PM2.5) pollution levels in Thailand. It aims to consolidate PM2.5 data and identify dominant features for predicting air quality index (AQI) values. The document outlines data sources on PM2.5 levels and weather parameters from 2016-2019 in various Thai regions. It proposes using support vector regression instead of linear regression to determine the most important features for AQI prediction, and shows SVR results in predicting actual PM2.5 data more accurately than linear regression.