Machine learning analysis of factors inducing the usage in public parks

Abstract

An online survey was conducted amongst residents in Koto ward, Tokyo, to investigate the factors inducing the usage in public parks by machine learning analysis. As a result, we found that the diversity in activities could binarize into two clusters: no-use cluster and use cluster, which was carried out by latent class analysis. Furthermore, the analyses revealed that the interaction between the family in green spaces and the frequency of use, which occurred in green ways, were the main factors inducing behavioral modification of activities in public parks by decision tree analysis.