Semi-Supervised Machine Learning Method for Predicting Observed Individual Risk Preference Using Gallup Data

Risk and uncertainty are vital for making economic decision. An individual's propensity to make riskier decisions depends on various circumstances. Most studies on general and financial risk preference focus on analyzing the correlation between the sociodemographic and economic factors, using ordinary least-square and supervised machine learning. This article aims to investigate the effects of social and economic covariates on an individual's willingness to take general risks (observed) using the GPS and Gallup datasets. Based on the available observed risk-taking data for one year, we propose a semi-supervised machine learning-based approach that can efficiently predict the observed risk index for those countries/individuals for years when the observed risk-taking index was not collected. We find that linear models are insufficient to capture certain patterns, and non-linear models, such as random forest regression, can obtain lower error than those reported in past literature. We also find that subjective well-being influences risk-taking behavior