bibtype B - Monography
ARLID 0524074
utime 20240111141036.5
mtime 20200504235959.9
ISBN 978-80-7378-404-1
title (primary) (eng) Discrete Compositional Models for Data Mining
publisher
place Praha
name MatfyzPress
pub_time 2019
specification
page_count 180 s.
media_type P
keyword compositional models
keyword data mining
keyword software
author (primary)
ARLID cav_un_auth*0101118
full_dept Department of Decision Making Theory
share 50
name1 Jiroušek
name2 Radim
institution UTIA-B
full_dept (cz) Matematická teorie rozhodování
full_dept (eng) Department of Decision Making Theory
department (cz) MTR
department (eng) MTR
fullinstit Ústav teorie informace a automatizace AV ČR, v. v. i.
author
ARLID cav_un_auth*0216188
full_dept Department of Decision Making Theory
share 50
name1 Kratochvíl
name2 Václav
institution UTIA-B
full_dept (cz) Matematická teorie rozhodování
full_dept Department of Decision Making Theory
department (cz) MTR
department MTR
country CZ
fullinstit Ústav teorie informace a automatizace AV ČR, v. v. i.
source
source_type pdf
url http://library.utia.cas.cz/separaty/2020/MTR/kratochvil-0524074.pdf
cas_special
project
ARLID cav_un_auth*0356801
project_id MOST-18-04
agency AV ČR
country CZ
country TW
abstract (eng) This brochure has been written with the support of the bilateral Czech-Taiwanese project Compositional models for data mining financially supported by the Ministry of Science and Technology, Taiwan, and by the Czech Academy of Sciences under Grant No. MOST-18-04 . The main output of the project, realized in 2018 and 2019, is a new supervised web system enabling researchers to learn probabilistic (compositional) models (both causal and stochastic) from data. We have opted for the web architecture for two reasons. First, we assume the system will be expanded in subsequent years, and the web application means that the system administrator only has to keep updated one version of program codes. Second, the system is accessible from any place in the world, so it can be applied not only by the members of research teams collaborating within the above-mentioned project but also by all interested researchers from anywhere inthe world. This book should serve as a manual for users of the data mining system. Nevertheless, since the system is based on the theory of compositional models, and no comprehensive text on this theory exists, we decided to set up this text from two parts. The first one describes the theoretical background on which the models constructed from data are based. It also includes chapters showing how the compositional models can be applied to data mining tasks. For this reason, the first part summarizes results scattered in a number of research journal and conference papers, mainly by R. Jiroušek and his coauthors Vl. Bína and V. Kratochvíl. This part, after introducing the notation from general probability theory, puts a special emphasis on the notion of stochastic (conditional) independence, without which one cannot distill knowledge from probability models. Chapters 2-5 sum up excerpts from the original research conference and journal papers. The importance of this part can be seen not only in the fact that it is the first time when these results are surveyed in one comprehensive text but also that it is presented using a new unifying notation, without which it might be difficult to see the links interconnecting individual parts of this theoretical approach.\n
RIV IN
FORD0 10000
FORD1 10200
FORD2 10201
reportyear 2021
num_of_auth 2
inst_support RVO:67985556
permalink http://hdl.handle.net/11104/0308418
confidential S
arlyear 2019
mrcbU10 2019
mrcbU10 Praha MatfyzPress
mrcbU12 978-80-7378-404-1
mrcbU14 SCOPUS
mrcbU24 PUBMED
mrcbU34 WOS
mrcbU56 pdf