Ústav teorie informace a automatizace

Jste zde

Publication details

Quantification of prior information revised

Journal Article

Kárný Miroslav, Khailova Natalia, Nedoma Petr, Böhm Josef

serial: International Journal of Adaptive Control and Signal Processing vol.15, 1 (2001), p. 65-84

research: AV0Z1075907

project(s): GA102/99/1564, GA ČR, GA102/99/1292, GA ČR, 1999/12058, IST

keywords: prior information, Bayesian estimation, prior design of adaptive controllers

preview: Download

abstract (eng):

Quantification of prior information about possibly high-dimensional unknown parameters of dynamic probabilistic models is adressed. Their prior probability density function (pdf) is chosen in conjugate form. Individual pieces of information are converted into a common basic of fictious data so that different nature and uncertainty levels are respected.

Cosati: 12B, 09I


21.12.2012 - 16:10