A Graph-Partitioning Based Continuous Optimization Approach to Semi-supervised Clustering Problems

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Liu, Wei, Liu, Xin, Ng, Michael K., Zhang, Zaikun
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910861539409920
author Liu, Wei
Liu, Xin
Ng, Michael K.
Zhang, Zaikun
author_facet Liu, Wei
Liu, Xin
Ng, Michael K.
Zhang, Zaikun
contents Semi-supervised clustering is a basic problem in various applications. Most existing methods require knowledge of the ideal cluster number, which is often difficult to obtain in practice. Besides, satisfying the must-link constraints is another major challenge for these methods. In this work, we view the semi-supervised clustering task as a partitioning problem on a graph associated with the given dataset, where the similarity matrix includes a scaling parameter to reflect the must-link constraints. Utilizing a relaxation technique, we formulate the graph partitioning problem into a continuous optimization model that does not require the exact cluster number, but only an overestimate of it. We then propose a block coordinate descent algorithm to efficiently solve this model, and establish its convergence result. Based on the obtained solution, we can construct the clusters that theoretically meet the must-link constraints under mild assumptions. Furthermore, we verify the effectiveness and efficiency of our proposed method through comprehensive numerical experiments.
format Preprint
id arxiv_https___arxiv_org_abs_2503_04447
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Graph-Partitioning Based Continuous Optimization Approach to Semi-supervised Clustering Problems
Liu, Wei
Liu, Xin
Ng, Michael K.
Zhang, Zaikun
Optimization and Control
Machine Learning
Semi-supervised clustering is a basic problem in various applications. Most existing methods require knowledge of the ideal cluster number, which is often difficult to obtain in practice. Besides, satisfying the must-link constraints is another major challenge for these methods. In this work, we view the semi-supervised clustering task as a partitioning problem on a graph associated with the given dataset, where the similarity matrix includes a scaling parameter to reflect the must-link constraints. Utilizing a relaxation technique, we formulate the graph partitioning problem into a continuous optimization model that does not require the exact cluster number, but only an overestimate of it. We then propose a block coordinate descent algorithm to efficiently solve this model, and establish its convergence result. Based on the obtained solution, we can construct the clusters that theoretically meet the must-link constraints under mild assumptions. Furthermore, we verify the effectiveness and efficiency of our proposed method through comprehensive numerical experiments.
title A Graph-Partitioning Based Continuous Optimization Approach to Semi-supervised Clustering Problems
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2503.04447