Papers
Topics
Authors
Recent
Search
2000 character limit reached

Exact Recovery of Clusters in Finite Metric Spaces Using Oracle Queries

Published 31 Jan 2021 in cs.LG and stat.ML | (2102.00504v2)

Abstract: We investigate the problem of exact cluster recovery using oracle queries. Previous results show that clusters in Euclidean spaces that are convex and separated with a margin can be reconstructed exactly using only O(log⁡n)O(\log n) same-cluster queries, where nn is the number of input points. In this work, we study this problem in the more challenging non-convex setting. We introduce a structural characterization of clusters, called (β,γ)(\beta,\gamma)-convexity, that can be applied to any finite set of points equipped with a metric (or even a semimetric, as the triangle inequality is not needed). Using (β,γ)(\beta,\gamma)-convexity, we can translate natural density properties of clusters (which include, for instance, clusters that are strongly non-convex in Rd\mathbb{R}^d) into a graph-theoretic notion of convexity. By exploiting this convexity notion, we design a deterministic algorithm that recovers (β,γ)(\beta,\gamma)-convex clusters using O(k2log⁡n+k2(6/βγ)dens(X))O(k^2 \log n + k^2 (6/\beta\gamma)^{dens(X)}) same-cluster queries, where kk is the number of clusters and dens(X)dens(X) is the density dimension of the semimetric. We show that an exponential dependence on the density dimension is necessary, and we also show that, if we are allowed to make O(k2+klog⁡n)O(k^2 + k\log n) additional queries to a "cluster separation" oracle, then we can recover clusters that have different and arbitrary scales, even when the scale of each cluster is unknown.

Citations (16)

Summary

No one has generated a summary of this paper yet.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Continue Learning

We haven't generated follow-up questions for this paper yet.