Deakin University
Browse

File(s) under permanent embargo

Multilevel clustering via wasserstein means

conference contribution
posted on 2017-01-01, 00:00 authored by N Ho, X L Nguyen, M Yurochkin, H H Bui, Viet Huynh, Quoc-Dinh Phung
Copyright 2017 by the author(s). We propose a novel approach to the problem of multilevel clustering, which aims to simultaneously partition data in each group and discover grouping patterns among groups in a potentially large hierarchically structured corpus of data. Our method involves a joint optimization formulation over several spaces of discrete probability measures, which are endowed with Wasserstein distance metrics. We propose a number of variants of this problem, which admit fast optimization algorithms, by exploiting the connection to the problem of finding Wasserstein barycenters. Consistency properties are established for the estimates of both local and global clusters. Finally, experiment results with both synthetic and real data arc presented to demonstrate the flexibility and scalability of the proposed approach.

History

Event

Machine Learning. International Conference (34th : 2017 : Sydney, New South Wales)

Volume

70

Pagination

2363 - 2378

Publisher

ICML

Location

Sydney, New South Wales

Place of publication

[Sydney, N.S.W.]

Start date

2017-08-06

End date

2017-08-11

ISBN-13

9781510855144

Language

eng

Publication classification

E Conference publication; E1 Full written paper - refereed

Title of proceedings

ICML 2017: Proceedings of the 34th International Conference in Machine Learning