Bayesian Online Multilabel Classification (BOMC)

version 1.0

September 29, 2011

Overview

BOMC is an open source toolkit for online multilabel classification using Bayesian models [1, 2]. It is implemented in F# 1.9.3.4 on Microsoft Visual Studio 2008, and can be compiled and run on Linux systems via Mono. The graphical model, as shown in Figure 1~3 of [1], is extend from TrueSkill^TM [2] to deal with multilabel, and the inference engine is expectation propagation.

We refer the interested users to two other illustrative F# implementations of TrueSkill^TM : original [2] and temporal [3].

Download

BOMC version 1.0

Disclaimer

BOMC is licensed under Mozilla Public License version 1.1. The authors are not responsible for any implications from the use of the software.

Contacts

Xinhua Zhang | Thore Graepel | Ralf Herbrich

References

[1]

Xinhua Zhang, Thore Graepel, Ralf Herbrich

Bayesian Online Learning for Multi-label and Multi-variate Performance Measures

International Conference on Artificial Intelligence and Statistics (AISTATS), 2010. [PDF]

[2]

Ralf Herbrich, Tom Minka, Thore Graepel
Trueskill^TM: A Bayesian skill ranking system.
Neural Information Processing Systems (NIPS) 2007.

[3]

Pierre Dangauthier, Ralf Herbrich, Tom Minka, and Thore Graepel
Trueskill^TM: Through Time: Revisiting the History of Chess.
Neural Information Processing Systems (NIPS) 2008.

Last modified: 29 September, 2011