Bayesian Learning With Unbounded Capacity From Heterogeneous And Set-Valued Data

Abstract

Large-scale and modern datasets have reshaped machine learning research and practices. They are notonly bigger in size, but predominantly heterogeneous and growing in their complexity. This proposalaims to advance machine learning methods grounded in the theory of recent Bayesian nonparametrics todeal with growing complexity and heterogeneity of large-scale data. The proposal is unique in its approachto deliver three new bodies of theory and techniques for: a) Bayesian nonparametric methods thatcan express and inference from heterogeneous, set-valued data sources with infinite model capacity, b)new framework for deterministic fast inference based on small-variance asymptotic analysis (SVAA) andWasserstein geometry and c) new applications in pervasive healthcare and exploiting electronic medicalrecords (EMR) data.

Document Details

Document Type
DoD Grant Award
Publication Date
Dec 05, 2016
Source ID
FA23861614138

Entities

People

  • Dinh Phung

Organizations

  • Air Force Office of Scientific Research
  • Deakin University
  • United States Air Force

Tags

Fields of Study

  • Computer science

Readers

  • Data Mining and Knowledge Discovery.
  • Distributed Systems and Data Platform Development
  • Graph Algorithms and Convex Optimization.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Microelectronics