Jump to contentJump to search


Anticlustering is the less well-known twin of the popular clustering method, which is ubiquitous in unsupervised learning. While cluster analysis is used to partition objects into homogeneous and well separated groups, anticlustering is used to divide objects into groups that are similar to each other. An example of an anticlustering problem is the assignment of test questions to different versions of an examination, to ensure the same level of difficulty for different cohorts of students. Anticlustering has important applications in a variety of research areas, including machine learning, artificial intelligence, psychology, operations research and network systems.

We developed the open source R package anticlust which makes the anticlustering methodology freely accessible (Papenberg and Klau, 2021). We are currently working on improved algorithms for anticlustering and are investigating novel applications of the anticlustering methodology. Contact  if you have questions on the anticlustering method or the anticlust package.

Responsible for the content: