"52"^^ . . "52"^^ . "IAA2075606" . . . . "0"^^ . "Adaptivn\u00ED dynamick\u00E9 prvky a jejich spojov\u00E1n\u00ED pro \u0159e\u0161en\u00ED dynamick\u00FDch \u00FAloh rozhodov\u00E1n\u00ED za neur\u010Ditosti" . . "http://www.isvav.cz/projectDetail.do?rowId=IAA2075606"^^ . "The project will contribute to development of design methods suitable for dynamic decision making. Adaptive systems, pattern recognition, neural networks and complexity theory lie in the project background. A design of %22simple%22 adaptive dynamic decision-making elements and their interconnections approximating the optimal strategy will serve as the central test-case. The project mile-stones are:- A novel fully probabilistic design of adaptive network elements promising their dual and robust nature.- Exploitation of adaptive elements in networks. This insufficiently exploited area promises a significant increase of flexibility and power of approximants.- The intended inter-disciplinary application of a variety of approximate methods of design and analysis to a common problem of decision-making under uncertainty is expected to provide a range of theoretical results related to the central project aim."@en . . . . . . . "1"^^ . "Adaptive dynamic elements and their interconnections for dynamic decision making under uncertainty"@en . . "0"^^ . . . . "Projekt p\u0159isp\u011Bje ke sjednocen\u00ED metod konstrukce realizovateln\u00FDch p\u0159ibl\u00ED\u017Een\u00ED vhodn\u00FDch pro \u0159e\u0161en\u00ED dynamick\u00FDch rozhodovac\u00EDch \u00FAloh. Vych\u00E1z\u00ED z teori\u00ED a postup\u016F adaptivn\u00EDch syst\u00E9m\u016F, rozpozn\u00E1v\u00E1n\u00ED obrazc\u016F, neuronov\u00FDch s\u00EDt\u00ED a teorie slo\u017Eitosti. \u00DAst\u0159edn\u00EDm %22testovac\u00EDm%22 probl\u00E9mem je n\u00E1vrh %22jednoduch\u00FDch%22 adaptivn\u00EDch dynamick\u00FDch rozhodovac\u00EDch prvk\u016F a zp\u016Fsob\u016F jejich kombinace do s\u00EDt\u011B aproximuj\u00EDc\u00ED optim\u00E1ln\u00ED strategii. Kl\u00ED\u010Dov\u00E9 body jsou: N\u00E1vrh adaptivn\u00EDch stavebn\u00EDch prvk\u016F s\u00EDt\u011B na z\u00E1klad\u011B p\u016Fvodn\u00ED pravd\u011Bpodobnostn\u00ED metodiky slibuj\u00EDc\u00ED jejich du\u00E1ln\u00ED a robustn\u00ED charakter. U\u017Eit\u00ED adaptivn\u00EDch dynamick\u00FDch prvk\u016F v s\u00EDt\u00EDch. Tato dosud m\u00E1lo prozkouman\u00E1 oblast slibuje v\u00FDrazn\u00E9 zv\u00FD\u0161en\u00ED flexibility aproximant\u016F.Kombinace r\u016Fzn\u00FDch p\u0159ibli\u017En\u00FDch metod n\u00E1vrhu a anal\u00FDzy na spole\u010Dn\u00FD probl\u00E9m dynamick\u00E9ho rozhodoov\u00E1n\u00ED za neur\u010Ditosti bude hlavn\u00EDm zdrojem mezioborov\u011B stimulovan\u00FDch teoretick\u00FDch poznatk\u016F sm\u011B\u0159uj\u00EDc\u00EDch k z\u00E1kladn\u00EDmu c\u00EDli projektu." . . "Byla navr\u017Eena, publikov\u00E1na a prakticky ov\u011B\u0159ena obecn\u00E1 konstrukce adaptivn\u00EDho p\u0159edpov\u00EDd\u00E1n\u00ED a \u0159\u00EDzen\u00ED pomoc\u00ED kooperuj\u00EDc\u00EDch jednoduch\u00FDch prvk\u016F. Projekt z\u00FA\u017Eil bari\u00E9ru mezi klasick\u00FDm rozhodov\u00E1n\u00EDm za neur\u010Ditosti a aproxima\u010Dn\u011B efektivn\u00EDmi neuronov\u00FDmi s\u00EDt\u011Bmi."@cs .