Variational relevance vector machine for tabular data

Dmitry Kropotov, Dmitry Vetrov, Lior Wolf, Tal Hassner

Research output: Contribution to journalConference articlepeer-review


We adopt the Relevance Vector Machine (RVM) framework to handle cases of tablestructured data such as image blocks and image descriptors. This is achieved by coupling the regularization coefficients of rows and columns of features. We present two variants of this new gridRVM framework, based on the way in which the regularization coefficients of the rows and columns are combined. Appropriate variational optimization algorithms are derived for inference within this framework. The consequent reduction in the number of parameters from the product of the table's dimensions to the sum of its dimensions allows for better performance in the face of small training sets, resulting in improved resistance to overfitting, as well as providing better interpretation of results. These properties are demonstrated on synthetic data-sets as well as on a modern and challenging visual identification benchmark.

Original languageEnglish
Pages (from-to)79-94
Number of pages16
JournalJournal of Machine Learning Research
StatePublished - 2010
Event2nd Asian Conference on Machine Learning, ACML 2010 - Tokyo, Japan
Duration: 8 Nov 201010 Nov 2010


  • Automatic relevance determination
  • Bayesian learning
  • Feature selection
  • Image classification
  • Relevance vector machine
  • Variational inference


Dive into the research topics of 'Variational relevance vector machine for tabular data'. Together they form a unique fingerprint.

Cite this