Editorial: Journal of Classification Vol. 40-2

In the first article, Temple considers the important topic of distance metrics. This article is published posthumously, and I am most grateful to Richard Bateman and his colleagues for helping shepherd it through the last stages of the review process. In the second contribution, Di Mari, Ingrassia and Punzo extend deviance measures to mixtures of generalized linear models. They do so locally, i.e., component-wise, and globally, leading to local and overall deviance R² measures. The third paper, by Furno, is a fascinating treatment of the important but often overlooked topic of computation of finite mixture estimators in the tails. In the fourth paper, Ahmed, Cassese, van Breukelen and Schepers introduce a new method — extended maximal interaction two-mode clustering (E-ReMI) — for studying two-way interaction in two-mode data. The fifth paper is a magnificent invited review article by Biernacki, Jacques and Keribin. The subject of the review is model-based co-clustering with a focus on high-dimension and estimation challenges. In the sixth paper of this issue, Patel and Vishwamitra introduce a spider monkey optimization algorithm for dynamic kernel clustering. The penultimate paper, by Kim, Oh and Lim, introduces a new method for clustering high-dimensional zero-inflated time series data based on a modified thick-pen transform. In the final article of this issue, Vicari and Giordani introduce a new approach, known as CPclus, for clusteirng three-way data. …