Models for Analyzing Factors and Classifying Informal Employment

This study develops a classification model for employment status in Thailand, categorized into formal and informal employment, using data from the 2024 Informal Employment Survey conducted by the National Statistical Office of Thailand. Two machine learning techniques, Random Forest and Logistic Regression, were applied, with five variable selection methods used for Logistic Regression. The Random Forest model provided the highest overall performance in classifying employment status. The variable importance analysis indicated that ‘‘type of wage received’’ and ‘‘industry sector’’ were the most influential factors in distinguishing informal workers. These findings were consistent with those selected by the Forward Selection method in Logistic Regression. Workers receiving monthly wages were significantly more likely to be formally employed compared to those receiving daily, weekly, or non-monetary compensation. Likewise, workers in the manufacturing and service/trade sectors were less likely to be informally employed compared to those in agriculture. These results enhance understanding of the socio-economic characteristics in Thailand associated with informal employment and support evidence-based policy formulation.