Cluster Analysis in Data‐Driven Management and Decisions

Abstract

Clustering plays an important role in management and decision‐making processes. This paper first discusses three types of cluster analysis methods—centroid‐based, connectivity‐based, and density‐based. Then the challenges to traditional clustering in new business environments are highlighted, with algorithmic extensions and innovative efforts for coping with data that is dynamic, large‐scale, representative, non‐convex, and consensus in nature. In addition, three application cases are illustrated, where clustering is incorporated into the overall solution in the contexts of management support, business of sharing economy, and healthcare decision assistance.