RUS  ENG
Full version
JOURNALS // Informatics and Automation // Archive

Tr. SPIIRAN, 2012 Issue 22, Pages 189–204 (Mi trspy526)

Bayesian rating systems with additional information on tournament results

S. I. Nikolenkoab, D. V. Serdyukb, A. V. Sirotkinbc

a St. Petersburg Department of V. A. Steklov Institute of Mathematics, Russian Academy of Sciences
b St. Petersburg Academic University — Nanotechnology Research and Education Centre of the Russian Academy of Sciences (the Academic University)
c St. Petersburg Institute for Informatics and Automation of RAS

Abstract: We consider the problem of constructing a Bayesian rating system that learns a general ranking function from the results of individual matches and/or tournaments (rankings on small subsets) Extending the results of [9], we construct a rating system that can take into account additional numerical information on tournament results. The paper contains detailed descriptions of Bayesian rating models, inference algorithms, and experimental results.

Keywords: Bayesian rating systems, probabilistic graphical models, factor graph, Bayesian inference.

UDC: 004.8

Received: 18.06.2012



© Steklov Math. Inst. of RAS, 2024