Computational Information Geometry | Frank NIELSEN
Computational Information Geometry
The field of computational information geometry (also named discrete information geometry) is interested in exploring the following three domains:
- Generic algorithms: Design meta-algorithms that can handle any arbitrary parameterized distance or loss functions.
(The usual class of parameterized distances are Bregman, Csiszar and Burbea-Rao divergences.)
Examples: Kmeans, EM, Voronoi diagrams, barycenters, smallest enclosing balls, ball trees, etc.
- Geometry of information.
Examples: Dually flat spaces of exponential/mixture families (VC-dimension of balls remains unchanged to d+1), Riemmanian geometry, etc.
- Novel applications.
Examples: Statistics, machine learning, computational geometry
Participate to the LIX Colloquium ETVC08(free registration) that will also gather experts of information geometry.
Visit the blog
...(still) in preparation...
Online December 2007.
Last updated, August 2008.
(c) Frank NIELSEN, All rights reserved.