Editorial

The first paper to appear in the last issue of the year is by Beibei Zhang and Rong Chen and provides a new approach for clustering time series. Eschewing a parametric approach, they utilize the KolmogrovSmirnov statistic to measure the distance between two different time series. This is cleverly done by looking at the similarity between the serial dependence structures of each series. Then, after the similarity between all pairs of time series is computed, any standard clustering procedure that allows for the analysis of a proximity matrix can be used. With the current approach, the toolbox for clustering time series data continues to grow (see Euan, Ombao, and Ortega, 2018; Michael and Melnykov, 2016; Rahmanishamsi, Dolati, and Aghabozorgi, 2018), and I expect that this will continue to be in the case in the near future. Further, an open area of research will be the comparison of the time series methods with other methods for clustering longitudinal data, such as growth mixture modeling (see Ram and Grimm, 2009). Given that each approach the problem from such disparate principles, finding common connections and points of differences may aid in the understanding of each.