"Probabilistic methods" . . . "0"^^ . "IAA2075302" . . "0"^^ . . . . "Projekt \u0159e\u0161\u00ED probl\u00E9my aplikac\u00ED mnohorozm\u011Brn\u00FDch pravd\u011Bpodobnostn\u00EDch model\u016F ve vybran\u00FDch technick\u00FDch oborech, p\u0159edev\u0161\u00EDm v oblasti rozpozn\u00E1v\u00E1n\u00ED obraz\u016F a modelov\u00E1n\u00ED obrazov\u00E9 informace. Vzhledem k tomu, \u017Ee tyto modely v pln\u00E9 obecnosti budou v\u017Edy p\u0159esahovat mo\u017Enosti po\u010D\u00EDta\u010Dov\u00E9ho zpracov\u00E1n\u00ED, v\u00FDzkum se zam\u011B\u0159\u00ED na speci\u00E1ln\u00ED t\u0159\u00EDdu aproximuj\u00EDc\u00EDch distribuc\u00ED, kterou naz\u00FDv\u00E1me komponentov\u00E9 modely. Tyto modely jsou definov\u00E1ny omezen\u00FDm po\u010Dtem parametr\u016F. Rozli\u0161ujeme dv\u011B podt\u0159\u00EDdy model\u016F: \u0159et\u011Bzcov\u00E9 modely (typick\u00FDmi p\u0159\u00EDklady jsou grafick\u00E9 markovsk\u00E9 modely) a lanov\u00E9 modely (reprezentovan\u00E9 modely zalo\u017Een\u00FDmi na sm\u011Bs\u00EDch). Pro ob\u011B uva\u017Eovan\u00E9 podt\u0159\u00EDdy budou \u0159e\u0161eny probl\u00E9my t\u00FDkaj\u00EDc\u00ED se odhadu parametr\u016F, v\u00FDb\u011Bru optim\u00E1ln\u00EDch (suboptim\u00E1ln\u00EDch) model\u016F, synt\u00E9zou model\u016F a n\u00E1vrhem efektivn\u00EDch algoritm\u016F pro jejich vyu\u017Eit\u00ED. \u0158e\u0161en\u00ED t\u011Bchto probl\u00E9m\u016F vy\u017Eaduje netrivi\u00E1ln\u00ED teoretickou anal\u00FDzu vlastnost\u00ED uva\u017Eovan\u00FDch model\u016F." . . . "2007-02-27+01:00"^^ . "Multi-dimensional Compound Probabilistic Models"@en . "http://www.isvav.cz/projectDetail.do?rowId=IAA2075302"^^ . . "Within the project number of original theoretical results on multidimensional efficiently reprezentable probabilistic models were deduced. These results supported development of computational algorithms and design of image and text processing procedures."@en . . " Bayesian networks" . . "The project deals with multi-dimensional probabilistic models and their application in several technical fields, mainly in pattern recognition and image modelling. Regarding the fact that these models, in a general from, will always be computationally intractable, the research focuses on special class of distributions,called compound models, serving as appropriate approximations. The approximating distributions are composed of several components, each of wich is defined with a reasonable number of parameters. We are distinguishing two main families of compound models: chain models (with typical representatives grtaphical Markov models) and hawser models (represented by finite mixture models). For both these subclasses of models parameter estimation, optimal (suboptimal) model selection, models synthesis along with efficient algorithms development for model construction will be proposed. The solution of these problems requires non-trivial theoretical model analysis."@en . "86"^^ . . " pattern recognition" . "86"^^ . " soft computing" . "Probabilistic methods; Bayesian networks; mixture-based distributions; composition of distributions; soft computing; applications; pattern recognition; model synthesis"@en . " applications" . "2013-06-28+02:00"^^ . . " mixture-based distributions" . "Multidimenzion\u00E1ln\u00ED komponentov\u00E9 pravd\u011Bpodobnostn\u00ED modely" . . "2007-12-01+01:00"^^ . . . . "2003-01-01+01:00"^^ . " composition of distributions" . . "Byla odvozena \u0159ada origin\u00E1ln\u00EDch teoretick\u00FDch poznatk\u016F o mnohodimenzion\u00E1ln\u00EDch efektivn\u011B reprezentovateln\u00FDch pravd\u011Bpodobnostn\u00EDch modelech. Tyto v\u00FDsledky byly pou\u017Eity k n\u00E1vrhu v\u00FDpo\u010Detn\u00EDch algoritm\u016F a k odvozen\u00ED nov\u00FDch metod pro zpracov\u00E1n\u00ED text\u016F a obr\u00E1zk\u016F."@cs . . . "1"^^ .