bibtype J - Journal Article
ARLID 0644070
utime 20260120122203.7
mtime 20260106235959.9
SCOPUS 105024112610
WOS 001643462100008
DOI 10.1109/LSP.2025.3640068
title (primary) (eng) Occam's Razor in Pooling of Probability Densities
specification
page_count 5 s.
media_type P
serial
ARLID cav_un_epca*0253212
ISSN 1070-9908
title IEEE Signal Processing Letters
volume_id 33
volume 1 (2026)
page_num 156-160
publisher
name Institute of Electrical and Electronics Engineers
keyword Information entropy
keyword Minimum relative entropy principle
keyword Probability density function
keyword Forgetting
author (primary)
ARLID cav_un_auth*0101124
name1 Kárný
name2 Miroslav
institution UTIA-B
full_dept (cz) Adaptivní systémy
full_dept (eng) Department of Adaptive Systems
department (cz) AS
department (eng) AS
full_dept Department of Adaptive Systems
fullinstit Ústav teorie informace a automatizace AV ČR, v. v. i.
source
url https://library.utia.cas.cz/separaty/2026/AS/karny-0644070.pdf
cas_special
project
project_id CA21169
agency EU-COST
country XE
ARLID cav_un_auth*0452289
abstract (eng) Geometric and linear poolings often serve for the fusion of the knowledge contained in a finite set of probability densities. Their pros and cons are relatively well understood. Many other ways have also been studied. A recent insightful survey letter by Koliander et al. inspects a range of pooling ways based on various axioms, optimisation and supra-Bayesian handling. The gained extensive option set makes the proper choice of the pooling function harder. This letter reduces the extent of unjustified options. It provides the optimisation-based selection among available options. Its steps are justified by well-established, axiomatically supported, minimum relative entropy and approximation principles. The text applies Occam’s razor to its theoretical tools, too. It simplifies the user’s choice of the pooling function and its weights. This weakens the possibility of a bad choice and opens the way to a range of applications.
result_subspec WOS
RIV BB
FORD0 10000
FORD1 10200
FORD2 10201
reportyear 2026
num_of_auth 1
inst_support RVO:67985556
permalink https://hdl.handle.net/11104/0374435
confidential S
mrcbC91 C
mrcbT16-e ENGINEERING.ELECTRICAL&ELECTRONIC
mrcbT16-f 3.9
mrcbT16-g 0.6
mrcbT16-h 6.1
mrcbT16-i 0.01691
mrcbT16-j 0.983
mrcbT16-k 18076
mrcbT16-q 167
mrcbT16-s 0.938
mrcbT16-y 29.89
mrcbT16-x 4.63
mrcbT16-3 6705
mrcbT16-4 Q1
mrcbT16-5 3.300
mrcbT16-6 635
mrcbT16-7 Q2
mrcbT16-C 68.4
mrcbT16-M 0.87
mrcbT16-N Q2
mrcbT16-P 68.4
arlyear 2026
mrcbU14 105024112610 SCOPUS
mrcbU24 PUBMED
mrcbU34 001643462100008 WOS
mrcbU63 cav_un_epca*0253212 IEEE Signal Processing Letters 33 1 2026 156 160 1070-9908 1558-2361 Institute of Electrical and Electronics Engineers