Are we living in a happy country: An analysis of national happiness from machine learning perspective

Abstract

National happiness has been actively studied during last ten years. The factor of happiness could be different due to different human perspective. The factors used in this work include both physical needs and the mental needs of humanity such as educational factor. This work identified more than 90 features that can be used to predict the country happiness. Unfortunately, manually analyzing the features is difficult and needs a lot of resources. Due to numerous size of the features, it is unwise to rely on the prediction of national happiness by manual analysis. That process will result in the high cost of analysis. Therefore, this work used machine learning technique which is a Support Vector Machine to learn and predicts the country happiness. Dimensionality reduction is also done in this work. Using the information gain technique, the features can be reduced. This technique is chosen due to its ability to explore the interrelationships among a set of variables. The selected features are also evaluated using the SVM classifier. Using the data of 187 countries from the UN Development Project, this work is able to identify which factor needed to be improved by a certain country to increase the happiness of their citizens.