Analyzing and predicting global happiness index via integrated multilayer clustering and machine learning models

Area of Science:

• Social Sciences
• Computational Social Science
• Psychology

Background:

• Global happiness prediction is complex, requiring advanced analytical methods.
• Understanding happiness drivers is crucial for policy development and societal well-being.
• Existing models may not fully capture the nuanced patterns of national happiness.

Purpose of the Study:

• To develop and validate a novel predictive framework for global happiness.
• To identify the key determinants influencing happiness scores across nations.
• To categorize countries into distinct happiness groups for targeted analysis.

Main Methods:

• Integration of unsupervised (K-Means clustering) and supervised (Random Forests, XGBoost) machine learning techniques.
• Hierarchical analysis incorporating cluster assignments as features in ensemble models.
• Utilizing data from the World Happiness Report for empirical analysis.

Main Results:

• A novel framework significantly improved happiness prediction by approximately 12% R².
• Global happiness scores were effectively categorized into three distinct groups: high, medium, and low.
• Social support and Gross Domestic Product (GDP) were identified as the most significant drivers of happiness.

Conclusions:

• The proposed hierarchical machine learning approach enhances the accuracy of global happiness prediction.
• Social support and GDP are critical factors for national well-being.
• Findings offer valuable insights for policymakers aiming to improve public happiness and social progress.