Analysis of World Happiness Report Dataset Using Machine Learning Approaches

Abstract

happiness is a dream goal to be achieved by governments and individuals and it can be considered as a proper measure of social development progress. The purpose of this paper is to conduct a study on World happiness report dataset, to classify the most critical variables regarding the life happiness score. The strong evidence of the identified main features classified from the outcomes of applying the supervised machine learning approaches using the Neural Network training model and the OneR models in classifications and feature selection. The trained model used in predictions revealed the insights derived from applying the data analysis, where the study found out that the GDP per capita is the critical indicator of life happiness score as well as the health life expectancy is the second primary feature. Findings from study evaluated using different performance metrics such as accuracy and confusion matrix to prove the insights gained from the data.