Determination of Dissaving Risk against Life Expectancy for Elderly People Through Clustering Analysis Applied to Additional Anonymous Data

In order to detect if the economic activity of elderly people at age sixty-five or older has begun to decline, anonymous data (AD) are used to extract characteristic features. AD are provided by the National Survey of Family Income and Expenditure published by the Ministry of Internal Affairs and Communications. So far, a methodology to detect dissaving risk among extremely low income and savings were validated for the AD on 2004. However, it would be meaningful to investigate if this method would be applicable as well in other AD datasets. Therefore, in this paper, with the addition of AD on 1999, 2009, and 2014, a method to detect dissaving risk is reanalyzed. Firstly, clustering analysis based on standardized income and savings for AD on 1999, 2009, and 2014 produced similar tendency as AD on 2004. Then clustering analysis based on savings is applied to the clusters with greatly low income and savings. As a result, it has been shown that a method to detect the data determined as dissaving risk was also applicable in the additional AD datasets.