HUMAN PERSONALITY CLASSIFICATION USING SUPERVISED MACHINE LEARNING ALGORITHMS

Abstract

According to certain definitions, "personality" refers to a person’s distinctive ways of thinking, feeling, and behaving in a variety of situations. The personality can be used to identify the behavior patterns of a human. The goal of this research is to categorize human personalities based on their behaviors.This study used data that was collected as Secondary data targeting the main five personality types considering behavioral tendencies, namely the supervisor, the commander, the inspector, the doer, and the idealist. After the pre-processing, six classification algorithms were used: Support Vector Machine (SVM), Random Forest, Naïve Bayes, Logistic regression,Multilayer Perception (MLP), and Decision Tree. The result was validated using 10-fold cross-validation. Based on the result, the highest accuracy is obtained in SVM with anaccuracy value of 88.5%. The highest precision,recall, f-measure, and lower error rates are obtained by comparing the above six supervised machine learning algorithms.