Citation

Neoh, Hoo Thye (2025) Machine Learning Interpretation framework for generic needs in Marital Satisfaction. PhD thesis, Multimedia University.

Full text not available from this repository.

Abstract

Marital relationship quality is vital to happiness and the mental or physical health of couple and family. A highly satisfying marriage is highly resistant to union dissolution and family fragmentation. As a key metric of relational health, marital satisfaction functions not only as a barometer of couple well-being but also as an aggregate measure of societal sustainability. In the Malaysian context, healthy family units enable the achievement of multiple Sustainable Development Goals (SDGs 1, 2, 3, 4, 5, 8, 10, 11, 12, 16, and 17). Extensive research documents the deleterious socioeconomic consequences of divorce, particularly for women and children. Youth raised without the benefits of two-parent households demonstrate increased vulnerability to various social problems, while elevated divorce rates collectively threaten sustainable social development. Despite national progress, Malaysia continues to grapple with persistently high divorce rates, highlighting the urgent need for effective interventions. Understanding cross-cultural predictors for marital satisfaction is difficult. Current interventions remain limited by their symptom-focused approaches and cultural specificity, failing to address universally applicable solutions across Malaysia's diverse ethnic landscape. The root cause of the failure to resolve the divorce problem lies in the lack of a healthy, universally applicable cross-cultural model of marital relationships. This study employs interpretable machine learning techniques to analyze cross-cultural data from 33 countries (N=7,767), with the primary objective of developing a comprehensive model that elucidates the complex interplay between generic human needs and marital satisfaction. The research addresses existing inconsistencies in marital satisfaction predictors across machine learning studies by integrating diverse theoretical perspectives on happiness and relational satisfaction. A 3-question screening tool (Minimal Viable Model) paired with Logistic Regression achieves 86% accuracy, enabling scalable, low-burden assessments. The key contributions of this research include the development of an Interpretable Machine Learning Model for Generic Needs for Marital Satisfaction and the creation of the Marital Relationship Risk Control Test (MRRCT) framework, which represents a major advancement in human-centric AI applications for social science. The findings offer substantial practical implications for evidence-based couple decision-making, targeted government policy formulation, cross-cultural research on universal values, and marital education programs focused on divorce prevention.

Item Type:	Thesis (PhD)
Additional Information:	Call No.: Q325.5 .N46 2025
Uncontrolled Keywords:	Machine learning — Social aspects
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Divisions:	Faculty of Computing and Informatics (FCI)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	15 Apr 2026 09:00
Last Modified:	15 Apr 2026 09:00
URII:	http://shdl.mmu.edu.my/id/eprint/15707

Downloads

Edit (login required)