Poster (Invited)

FZJ-2025-02405

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png FairDen: Fair Density-Based Clustering

Krieger, L. (Corresponding author)FZJ* ; Beer, A. ; Matthews, P. ; Thiesson, A. ; Assent, I.FZJ*

2025

Abstract: Fairness in data mining tasks like clustering has recently become an increasinglyimportant aspect. However, few clustering algorithms exist that focus on fairgroupings of data with sensitive attributes. Including fairness in the clusteringobjective is especially hard for density-based clustering, as it does not directlyoptimize a closed form objective like centroid-based or spectral methods.This paper introduces FairDen, the first fair, density-based clustering algorithm.We capture the dataset’s density-connectivity structure in a similarity matrix thatwe manipulate to encourage a balanced clustering. In contrast to state-of-theart, FairDen inherently handles categorical attributes, noise, and data with severalsensitive attributes or groups. We show that FairDen finds meaningful and fairclusters in extensive experiments.

---

Contributing Institute(s):

1. Datenanalyse und Maschinenlernen (IAS-8)

Research Program(s):

1. 5112 - Cross-Domain Algorithms, Tools, Methods Labs (ATMLs) and Research Groups (POF4-511) (POF4-511)

Appears in the scientific report 2025

---