Comparative Study of Inference Methods for Bayesian Nonnegative Matrix Factorisation

In this paper, we study the trade-offs of different inference approaches for Bayesian matrix factorisation methods, which are commonly used for predicting missing values, and for finding patterns in the data. In particular, we consider Bayesian nonnegative variants of matrix factorisation and tri-factorisation, and compare non-probabilistic inference, Gibbs sampling, variational Bayesian inference, and a maximum-a-posteriori approach. The variational approach is new for the Bayesian nonnegative models. We compare their convergence, and robustness to noise and sparsity of the data, on both synthetic and real-world datasets. Furthermore, we extend the models with the Bayesian automatic relevance determination prior, allowing the models to perform automatic model selection, and demonstrate its efficiency. Code and data related to this chapter are availabe at: https://github.com/ThomasBrouwer/BNMTF_ARD.

• Publication year

  2017

• Venue

  ECML/PKDD

• Publication date

  2017-07-13

• Fields of study

  Mathematics, Computer Science

• Identifiers
  DOI 10.1007/978-3-319-71249-9_31arXiv 1707.05147
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Bayesian Hybrid Matrix Factorisation for Data Integration

  2017influential reference
• Harnessing Connectivity in a Large-Scale Small-Molecule Sensitivity Dataset.

  2015cited by this paper
• Markov Chain Monte Carlo and Variational Inference: Bridging the Gap

  2014cited by this paper
• Non-negative matrix factorization by maximizing correntropy for cancer clustering

  2013cited by this paper
• The Cancer Cell Line Encyclopedia enables predictive modeling of anticancer drug sensitivity

  2012cited by this paper
• Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic biomarker discovery in cancer cells

  2012cited by this paper
• Predicting drug-target interactions from chemical and genomic kernels using Bayesian matrix factorization

  2012cited by this paper
• Supplementary material to “ Automatic Relevance Determination in Nonnegative Matrix Factorization with the β-Divergence ”

  2012cited by this paper
• Automatic Relevance Determination in Nonnegative Matrix Factorization with the /spl beta/-Divergence

  2011cited by this paper
• Bayesian Group Factor Analysis

  2011cited by this paper
• Bayesian methods to detect dye‐labelled DNA oligonucleotides in multiplexed Raman spectra

  2011cited by this paper
• Bayesian CCA via Group Sparsity

  2011cited by this paper
• On Smoothing and Inference for Topic Models

  2009cited by this paper
• Probabilistic matrix tri-factorization

  2009influential reference
• Reversible Jump MCMC for Non-Negative Matrix Factorization

  2009cited by this paper
• Bayesian Non-negative Matrix Factorization

  2009influential reference
• A Non-negative Matrix Tri-factorization Approach to Sentiment Classification with Lexical Prior Knowledge

  2009cited by this paper
• Relation Prediction in Multi-Relational Domains using Matrix Factorization

  2008cited by this paper
• Collaborative Filtering Using Orthogonal Nonnegative Matrix Tri-factorization

  2007cited by this paper
• Probabilistic Matrix Factorization

  2007cited by this paper
• Orthogonal nonnegative matrix t-factorizations for clustering

  2006cited by this paper
• Two-Dimensional Non-negative Matrix Factorization for Face Representation and Recognition

  2005cited by this paper
• The variational Bayesian EM algorithm for incomplete data: with application to scoring graphical model structures

  2003cited by this paper
• Unsupervised Learning of Finite Mixture Models

  2002cited by this paper
• Algorithms for Non-negative Matrix Factorization

  2000cited by this paper
• Learning the parts of objects by non-negative matrix factorization

  1999cited by this paper

• Bayesian Nonnegative Tensor Completion With Automatic Rank Determination

  2025influential citation
• Bayesian non-negative matrix factorization with Student's t-distribution for outlier removal and data clustering

  2024cites this paper
• <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline" id="d1e4378" altimg="si182.svg"><mml:mi>β</mml:mi></mml:math>-divergence NMF with biorthogonal regularization for data representation

  2023cites this paper
• A Variational Bayesian Gaussian Mixture-Nonnegative Matrix Factorization Model to Extract Movement Primitives for Robust Control

  2022cites this paper
• Towards Robust Community Detection via Extreme Adversarial Attacks

  2022cites this paper
• Variational Bayesian Orthogonal Nonnegative Matrix Factorization Over the Stiefel Manifold

  2022cites this paper
• STRONG: metagenomics strain resolution on assembly graphs

  2021cites this paper
• Entropy Minimizing Matrix Factorization

  2021cites this paper
• Indian Buffet Process-Based on Nonnegative Matrix Factorization with Single Binary Component

  2021cites this paper
• Low-rank nonnegative matrix factorization on Stiefel manifold

  2020cites this paper
• Metagenomics Strain Resolution on Assembly Graphs

  2020cites this paper
• Learning Nonnegative Factors From Tensor Data: Probabilistic Modeling and Inference Algorithm

  2020cites this paper
• Bayesian Non-Negative Matrix Factorization With Adaptive Sparsity and Smoothness Prior

  2019cites this paper
• VARIATIONAL BAYESIAN PARTIALLY OBSERVED NON-NEGATIVE TENSOR FACTORIZATION

  2018influential citation
• Probabilistic Sparse Non-negative Matrix Factorization

  2018cites this paper
• Prior and Likelihood Choices for Bayesian Matrix Factorisation on Small Datasets

  2017cites this paper