Machine Learning Approaches to Predict Divorce-Related Single Motherhood

Abstract

This study aims to apply machine learning for determining the factors that may lead to divorce in single motherhood in Thailand. The research is based on the Cross-Industry Standard Process (CRISP) that consists of six steps from problem understanding to deployment using Python as a tool. Thailand's National Statistical Office showed the decline in marriage registration rates since 2004 while divorce rates raised. A questionnaire was designed to collect a sample of 38 female participants specified aged>18 years with a minimum of one child. The dataset was partitioned using a 75:25 ratio for training and testing purposes. Three machine learning algorithms were implemented: Random Forest, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). The results showed that KNN and SVM algorithms predict divorce leading to single motherhood with 100% accuracy. This preliminary study provides a methodological framework for investigating divorce prediction using machine learning, however the more comprehensive research incorporating sociocultural dimensions specific to Thailand.