VertCoHiRF: Decentralized Vertical Clustering Beyond k-means

TL;DR

VertCoHiRF introduces a fully decentralized vertical federated clustering framework that achieves hierarchical consensus without sharing feature-dependent data, enhancing privacy, robustness, and interpretability in multi-party settings.

Contribution

It proposes a novel decentralized clustering method that relies on identifier-level consensus and ordinal rankings, avoiding centralized coordination and feature sharing.

Findings

Competitive clustering performance demonstrated in experiments.

Supports overlapping feature partitions and heterogeneous local methods.

Provides an interpretable hierarchical clustering structure.

Abstract

Vertical Federated Learning (VFL) enables collaborative analysis across parties holding complementary feature views of the same samples, yet existing approaches are largely restricted to distributed variants of $k$-means, requiring centralized coordination or the exchange of feature-dependent numerical statistics, and exhibiting limited robustness under heterogeneous views or adversarial behavior. We introduce VertCoHiRF, a fully decentralized framework for vertical federated clustering based on structural consensus across heterogeneous views, allowing each agent to apply a base clustering method adapted to its local feature space in a peer-to-peer manner. Rather than exchanging feature-dependent statistics or relying on noise injection for privacy, agents cluster their local views independently and reconcile their proposals through identifier-level consensus. Consensus is achieved via decentralized ordinal ranking to select representative medoids, progressively inducing a shared hierarchical clustering across agents. Communication is limited to sample identifiers, cluster labels, and ordinal rankings, providing privacy by design while supporting overlapping feature partitions and heterogeneous local clustering methods, and yielding an interpretable shared Cluster Fusion Hierarchy (CFH) that captures cross-view agreement at multiple resolutions.We analyze communication complexity and robustness, and experiments demonstrate competitive clustering performance in vertical federated settings.