BIRS Workshop Lecture Videos
New algorithms for sampling closed and/or confined equilateral polygons Cantarella, Jason
This talk gives an overview of new Markov chain methods based on symplectic geometry for understanding and sampling spaces of fixed edgelength polygons. The new methods are simple to code, and allow you to sample polygons in several different confinement models, as well as to prove some theorems about closed and confined random polygons. These are the first Markov chain methods for closed and confined polygons with a proof of geometric convergence to the correct probability measure on polygon space and statistically consistent error estimators for Markov chain Monte Carlo integration. The talk describes joint work with Clayton Shonkwiler (UGA).
Item Citations and Data
Attribution-NonCommercial-NoDerivs 2.5 Canada