A study of statistical data to analyze community Inequality in Roi Et province with machine learning

Abstract

This article aimed to study: (1) the characteristics of community-level inequality in Roi Et Province, (2) the analysis of data using statistical methods and machine learning techniques to identify key factors influencing inequality, and (3) the proposal of systematic policies and local development guidelines. This study employed a mixed-methods approach, utilizing both primary data from household surveys and secondary data from relevant governmental agencies. Data collection instruments included household questionnaires and statistical records, while data analysis employed descriptive statistics, K-Means clustering, and decision tree modeling. The findings revealed that: (1) the sample group in Tha Muang Subdistrict, Selaphum District, had a low average monthly income (9,500 Baht) and perceived inequality at a high level, particularly regarding income and access to public services; (2) K-Means clustering classified the study areas into three levels of inequality: low, medium, and critical, with Kaset Wisai District identified as the most critical area due to having the highest number of poor households; and (3) decision tree modeling indicated that the “number of poor households” and “education” were the most influential variables in determining inequality levels. These findings provide an empirical basis for designing targeted policy interventions, such as improving education, developing infrastructure, and enhancing access to welfare, thereby fostering equity and long-term sustainable development.