Institute of Information Science, Academia Sinica

Events

Print

Press Ctrl+P to print from browser

Bayesian Inference of Rankings in Competition and Preference Networks

:::

Bayesian Inference of Rankings in Competition and Preference Networks

  • LecturerProf. Juyong Park (KAIST)
    Host: Sheng-Wei (Kuan-Ta) Chen
  • Time2013-02-21 (Thu.) 11:15 ~ 12:00
  • LocationAuditorium 106 at New IIS Building
Abstract

Competition and reward is a significant mechanism in the function and evolution of many complex systems. But determining the dominance hierarchy of the system's constituent parts from the strongest to the weakest, the basis for determining the reward, is often an ambiguous task due to the incomplete nature of the network representing the competitions. Here we introduce a method of inferring the natural rankings based on Bayes' theorem. We show analytically that the Bayesian update of the winning probability can be mapped into a single-parameter fitness model, yielding an exact expression for natural ranking and uncertainty.

BIO

Juyong Park is an assistant professor at Graduate School of Culture Technology in KAIST. Juyong received his doctoral degree is physics at the University of Michigan in 2006. His research interests are in network theory and its applications. He is the inventor of the Wolverine Ranking System for US college football, which allows him to join his hobby and work.