Determining the Happiness Class of Countries with Tree-Based Algorithms in Machine Learning

Abstract

Today, the concept of happiness is a frequently researched subject in the fields of economy, medicine, and social and political fields, as well as psychology. It has been an important research area for everyone, from policymakers to companies, to determine the factors affecting happiness. With machine learning algorithms, it is possible to make classifications with very high accuracy. The aim of this study is to use tree-based machine learning algorithms to classify the happiness scores of countries. In order to accomplish this, data from the World Happiness Index published in 2022 were used. On these data, tree-based algorithms CART, tree-based ensemble algorithms Bagging, and Random Forest were used. The test data of the model were obtained with 85% precision, recall, and F1 metrics, which were calculated using Bagging and Random Forest algorithms. The outcomes of the models obtained during the study were interpreted.