Neural Network Analysis of the Determinants of Happiness of Community Residents: Application of the Multi-Layer Perceptron Analysis Method


  • Young-Chool Choi
  • Youngwha Kee


This study is based on analysis of a questionnaire survey conducted in 2021 by the Center for Community Well-being Research at the Graduate School of Public Administration at Seoul National University in Korea, which targeted all local authority residents in Korea. The aim of this study is to identify the factors that affect the happiness level of local residents in Korea by means of an artificial neural network analysis. Few studies have been conducted that analyse the important factors which affect the happiness level of local residents in Korea using the artificial neural network multi-layer perceptron model. In this study, among artificial neural network analyses, relative importance analysis was conducted for independent variables, using the weight values ​​for each node calculated via the multi-layer perceptron analysis method. It was found that the important factors affecting happiness levels were marital status, age, period of residence, income level, area of residence (urban or rural), religion, occupation type, residence type, gender and education level.