Multi-view Deep Gaussian Process with a Pre-training Acceleration Technique

  • Han Zhu
  • , Jing Zhao*
  • , Shiliang Sun
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Deep Gaussian process (DGP) is one of the popular probabilistic modeling methods, which is powerful and widely used for function approximation and uncertainty estimation. However, the traditional DGP lacks consideration for multi-view cases in which data may come from different sources or be constructed by different types of features. In this paper, we propose a generalized multi-view DGP (MvDGP) to capture the characteristics of different views and model data in different views discriminately. In order to make the proposed model more efficient in training, we introduce a pre-training network in MvDGP and incorporate stochastic variational inference for fine-tuning. Experimental results on real-world data sets demonstrate that pre-trained MvDGP outperforms the state-of-the-art DGP models and deep neural networks, achieving higher computational efficiency than other DGP models.

Original languageEnglish
Title of host publicationAdvances in Knowledge Discovery and Data Mining - 24th Pacific-Asia Conference, PAKDD 2020, Proceedings
EditorsHady W. Lauw, Ee-Peng Lim, Raymond Chi-Wing Wong, Alexandros Ntoulas, See-Kiong Ng, Sinno Jialin Pan
PublisherSpringer
Pages299-311
Number of pages13
ISBN (Print)9783030474355
DOIs
StatePublished - 2020
Event24th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2020 - Singapore, Singapore
Duration: 11 May 202014 May 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12085 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference24th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2020
Country/TerritorySingapore
CitySingapore
Period11/05/2014/05/20

Keywords

  • Deep Gaussian process
  • Multi-view learning
  • Pre-training technique
  • Stochastic optimization
  • Variational inference

Fingerprint

Dive into the research topics of 'Multi-view Deep Gaussian Process with a Pre-training Acceleration Technique'. Together they form a unique fingerprint.

Cite this