Bayesian Multi-Hyperplane Machine for pattern recognition

Khanh Nguyen, Trung Le, Tu Dinh Nguyen, Dinh Phung

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearch

Abstract

Current existing multi-hyperplane machine approach deals with high-dimensional and complex datasets by approximating the input data region using a parametric mixture of hyperplanes. Consequently, this approach requires an excessively time-consuming parameter search to find the set of optimal hyper-parameters. Another serious drawback of this approach is that it is often suboptimal since the optimal choice for the hyper-parameter is likely to lie outside the searching space due to the space discretization step required in grid search. To address these challenges, we propose in this paper BAyesian Multi-hyperplane Machine (BAMM). Our approach departs from a Bayesian perspective, and aims to construct an alternative probabilistic view in such a way that its maximum-a-posteriori (MAP) estimation reduces exactly to the original optimization problem of a multi-hyperplane machine. This view allows us to endow prior distributions over hyper-parameters and augment auxiliary variables to efficiently infer model parameters and hyper-parameters via Markov chain Monte Carlo (MCMC) method. We then employ a Stochastic Gradient Descent (SGD) framework to scale our model up with ever-growing large datasets. Extensive experiments demonstrate the capability of our proposed method in learning the optimal model without using any parameter tuning, and in achieving comparable accuracies compared with the state-of-art baselines; in the meantime our model can seamlessly handle with large-scale datasets.

Original languageEnglish
Title of host publication2018 24th International Conference on Pattern Recognition (ICPR)
Subtitle of host publicationAug. 20 2018 to Aug. 24 2018 Beijing, China
EditorsCheng-Lin Liu, Rama Chellappa, Matti Pietikäinen
Place of PublicationPiscataway NJ USA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages609-614
Number of pages6
ISBN (Electronic)9781538637883, 9781538637876
ISBN (Print)97815386-37890
DOIs
Publication statusPublished - 2018
Externally publishedYes
EventInternational Conference on Pattern Recognition 2018 - Beijing, China
Duration: 20 Aug 201824 Aug 2018
Conference number: 24th
http://www.icpr2018.org/

Conference

ConferenceInternational Conference on Pattern Recognition 2018
Abbreviated titleICPR 2018
CountryChina
CityBeijing
Period20/08/1824/08/18
Internet address

Cite this