Application of machine learning to cluster countries based on dominant parameters of happiness index

Abstract

Background: Several researches have been done to study the role of various psychological factors to predict the happiness. The terms happiness, well-being, quality of life is used interchangeably. There is no universal definition of happiness, as happiness means different things to different people. It changes with time, age, place etc. Some of the countries have started measuring growth rate on the basis of happiness index of their nation’s population. Purpose: To explore the relationship among the various parameters of happiness index and designing a classification model for the various countries on the basis of their happiness. Materials and Methods: In this research work we have applied machine learning model on the datasets available for the year 2018 to 2020 to find the correlation between various global parameters adapted for identifying the impact of those parameters on the happiness index for the 156 countries. Results: With the help unsupervised K-means learning method, we can analyze the importance of various parameters and can put the country in the class/cluster like excellent, moderate, average or below average. The above-mentioned classes signify the placement of the country and accurate classification on basis of various parameters. Conclusions: This paper discusses the current factors effecting the happiness index of the country. The concept of happiness index and dimensions that effects the happiness index of the nations is always the matter of discussion and contradiction. There is no such comprehensive work done on the global parameters of happiness index in the machine learning field. Keywords: machine learning, classification, K-Means method, happiness, world record.