Probability-Based Settings for Knowledge Update and Belief Revision
|
Supervisor |
|
|
Suitable for |
Abstract
Various qualitative-logical settings for modeling knowledge update and belief revision in a multi-agent context have been proposed by logicians and computer-scientists. On the other hand, there are various quantitative-probabilistic settings, mostly based on Bayesian updating. The project is to investigate the relations between some of these two types of settings, and explore ways to combine them into a unified probabilistic-logical approach to belief change. Such a unified setting could be very useful for AI applications.
Background Needed
This project assumes some familiarity with some basic notions in Logic and in Probability Theory.