Graduate Computational Algebraic Geometry Seminar
Jon Yaggie
UIC
Conditional Probabilistic Knowledge Bases using Gröbner Bases
Abstract: Under the circumstances where conditional knowledge is associated
with rational probabilities, the maximum entropy principle can be used
to generate a set of polynomial equations.
These polynomial equations can then be used to check consistency and
infer probabilities for new information via Gröbner bases computations
(Kern-Isberner, Wilhelm, Beierle).
I will review these theoretic results and demonstrate
an implementation using Sage.
Tuesday February 9, 2016 at 4:00 PM in SEO 1227