A Comparison of Machine Learning Algorithms for Emotional Intelligence Measure

Abstract

Emotional Intelligence (EI) refers to a person's ability to recognize, express, understand, manage, and use his or her own and others' emotions. EI is currently regarded as a distinct course of action, capable of excelling in a variety of disciplines. It is now used to make fascinating and significant forecasts about a person's life outcomes. Automated emotional information processing has become a trend in recent research approaches, with Machine Learning (ML) development. ML can assist EI in light of the success it has experienced in several other fields. ML and the deep learning approach can be used for the prediction of certain human cognition aspects like intelligence, working memory, attention, or focus. This study is based on a comparison of different Machine Learning Algorithms (MLA) for regression analysis of PEC (Profile of Emotional Competence) data. The regression algorithms SVM (Support Vector Machine – Regression), k-NN (k-Nearest Neighbor), and Ridge Regression are applied to the dataset. These MLAs were implemented in Python, trained using 70% of the PEC dataset and tested with 30% data. A comparative analysis evaluates the performance of MLAs using error metrics like ME, RMSE, MAE, EVS, coefficient of determinant and model accuracy, on the PEC measure of EI. The result identified the performance of SVM and k-NN in different EI domains and Ridge Regression as the best performing method on the given PEC data. This work contributes to a better understanding of the ML approach to the PEC EI measure and opens up the possibility of using the same approach for other EI measures.