Predicting Personality Traits via Psychometric Assessments: Insights from Machine Learning

This study investigates the association between Ayurvedic prakriti and cognitive personality traits using machine learning (ML) models. A random sample of 406 subjects from the USA $(\mathbf{N}=204)$ and India $(\mathrm{N}=202)$ was used. To balance the dataset, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. Various ML methods, including Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Support Vector Classifier (SVC), and Nearest Neighbor (NN), were employed to predict personality based on prakriti, emotional quotient (EQ), intelligence quotient (IQ), and risk-taking behavior. The ensemble learning approach showed significant improvements over traditional methods. The NN algorithm achieved the highest accuracy of 0.65 on the SMOTE data. Classification metrics such as precision, recall, F score, and accuracy were evaluated, demonstrating that ensemble techniques enhance personality trait prediction accuracy.