Established in 2005 under support of MŠMT ČR (project 1M0572)

Lectures and Presetations

Bayesian Learning: Back To Basis.

From:
Apr. 11 2006 2:00PM
To:
Apr. 11 2006 3:00PM
Place:
místnost 474, ÚTIA AV ČR
Description:
Bayesian estimation is a standard paradigm of learning that fits well to a subsequent decision making. Never ending controversy between it and Fisherian formulation of statistics and its known restrictions make useful a periodic way back to its basis. In this talk, an attempt is made in this respect. It may serve as
- an outline of a possible set of assumptions hidden behind the use of Bayesian estimation
- a description of very basic Bayesian tools - sketch of main open learning problems

Warning: the talk is given by an engineer so that no mathematical rigor can be expected.
 
Copyright 2005 DAR XHTML CSS