Dynamic Algorithm for Explainable k-medians Clustering under lp Norm

TL;DR

This paper introduces the first general algorithm for explainable k-medians clustering under any lp norm, providing approximation guarantees and an efficient dynamic implementation for evolving datasets.

Contribution

It presents a novel algorithm for explainable k-medians clustering under all lp norms for the first time, with proven approximation bounds and a dynamic update method.

Findings

Achieves an O(p(log k)^{1 + 1/p - 1/p^2}) approximation for all p >= 1.

Improves bounds for p=2 to O(log^{3/2}k).

Matches tight bounds for p=1 up to a log log k factor.

Abstract

We study the problem of explainable k-medians clustering introduced by Dasgupta, Frost, Moshkovitz, and Rashtchian (2020). In this problem, the goal is to construct a threshold decision tree that partitions data into k clusters while minimizing the k-medians objective. These trees are interpretable because each internal node makes a simple decision by thresholding a single feature, allowing users to trace and understand how each point is assigned to a cluster. We present the first algorithm for explainable k-medians under lp norm for every finite p >= 1. Our algorithm achieves an O(p(log k)^{1 + 1/p - 1/p^2}) approximation to the optimal k-medians cost for any p >= 1. Previously, algorithms were known only for p = 1 and p = 2. For p = 2, our algorithm improves upon the existing bound of O(log^{3/2}k), and for p = 1, it matches the tight bound of log k + O(1) up to a multiplicative O(log log k) factor. We show how to implement our algorithm in a dynamic setting. The dynamic algorithm maintains an explainable clustering under a sequence of insertions and deletions, with amortized update time O(d log^3 k) and O(log k) recourse, making it suitable for large-scale and evolving datasets.