Published July 1994 | Version v1
Journal article

Bayesian probability theory and inverse problems

Creators

  • 1. Uniwersytet Jagiellonski, Inst. Fizyki, Cracow (Poland)

Description

Bayesian probability theory is applied to approximate solving of the inverse problems. In order to solve the moment problem with the noisy data, the entropic prior is used. The expressions for the solution and its error bounds are presented. When the noise level tends to zero, the Bayesian solution tends to the classic maximum entropy solution in the L2 norm. The way of using spline prior is also shown. (author)

Additional details

Publishing Information

Journal Title
Acta Physica Polonica. Series B
Journal Volume
B25
Journal Issue
7
Journal Page Range
p. 1099-1108.
ISSN
0587-4254
CODEN
APOBBB

INIS

Country of Publication
Poland
Country of Input or Organization
Poland
INIS RN
29002595
Subject category
S71: CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSICS;
Descriptors DEI
CALCULATION METHODS; DIFFERENTIAL EQUATIONS; ENTROPY; INTEGRAL EQUATIONS; MOMENTS METHOD
Descriptors DEC
EQUATIONS; PHYSICAL PROPERTIES; THERMODYNAMIC PROPERTIES

Optional Information

Contract/Grant/Project number
KBN Grant No 2P30207904
Notes
11 refs.
Funding organization
Polish Committee for Scientific Research, Warsaw (Poland).