Patrik Norén Aalto-universitetet: Algebraic Statistics

Abstract: Algebraic statistics is about trying to solve statistical problems by using machinery from modern algebra. Problems in statistics are close to problems in algebra as many statistical models are algebraic or semi-algebraic sets. Algebraic geometry provides powerful machinery to study these kinds of sets, in statistics this can be used to for example answer questions about identifiability. Commutative algebra also gives new methods to do Markov chain Monte Carlo simulations by using Gröbner bases. I will give an overview of the new methods and ideas arising from algebraic statistics.

Tillbaka till huvudsidan.