"data mining; evolutionary techniques; genetic algorithm; inductive modelling; optimization"@en . . "7"^^ . "In this paper methods inspired by nature for models optimizations are shown. We describe mainly a~novel approach based on a~standard genetic algorithm advanced by some new parameters of individuals, methods of reproduction and adaptation, a way of individuals' encoding, a population of individuals representation, and a population size balancing technique. This method called Continual Evolution Algorithm combines evolution based process and gradient methods of individuals adaptation. Whole process runs in a continual time using a variable size of population. Motivations for using and improving methods for creation and optimization of models are a better adaptation of the models describing some given data or systems and knowledge extraction from such models."@en . . "RIV/68407700:21230/07:03129973" . "Z(MSM6840770012)" . . "21230" . "In this paper methods inspired by nature for models optimizations are shown. We describe mainly a~novel approach based on a~standard genetic algorithm advanced by some new parameters of individuals, methods of reproduction and adaptation, a way of individuals' encoding, a population of individuals representation, and a population size balancing technique. This method called Continual Evolution Algorithm combines evolution based process and gradient methods of individuals adaptation. Whole process runs in a continual time using a variable size of population. Motivations for using and improving methods for creation and optimization of models are a better adaptation of the models describing some given data or systems and knowledge extraction from such models." . "2"^^ . . . "2007-04-24+02:00"^^ . "2"^^ . "RIV/68407700:21230/07:03129973!RIV08-MSM-21230___" . "978-80-86840-30-7" . "Ro\u017Enov pod Radho\u0161t\u011Bm" . "Proceedings of 41th Spring International Conference MOSIS\\'07" . "436329" . "P\u0159\u00EDrodou inspirovan\u00E9 metody pro vytv\u00E1\u0159en\u00ED a optimalizaci model\u016F"@cs . . "P\u0159\u00EDrodou inspirovan\u00E9 metody pro vytv\u00E1\u0159en\u00ED a optimalizaci model\u016F"@cs . "MARQ" . . . . . "56;62" . "[08F7DE005B15]" . . . . . . "Nature Inspired Methods for Models Creation and Optimization" . "Nature Inspired Methods for Models Creation and Optimization"@en . "Ostrava" . "Buk, Zden\u011Bk" . . "\u0160norek, Miroslav" . "Nature Inspired Methods for Models Creation and Optimization" . "Tento \u010Dl\u00E1nek popisuje p\u0159\u00EDrodou inspirovan\u00E9 metody pro optimalizaci model\u016F. Pop\u00ED\u0161eme novou metodu zalo\u017Eenou na standardn\u00EDm genetick\u00E9m algoritmu roz\u0161\u00ED\u0159en\u00E9m o n\u011Bkter\u00E9 nov\u00E9 parametry jedinc\u016F, metody reprodukce a adaptace, k\u00F3dov\u00E1n\u00ED jedinc\u016F, reprezentaci populace a techniky \u0159\u00EDzen\u00ED velikosti populace. Tato metoda - Algoritmus kontinu\u00E1ln\u00ED evoluce (Continual Evolution Algorithm) kombinuje evolu\u010Dn\u00ED p\u0159\u00EDstup a gradientn\u00ED metody adaptace jedinc\u016F. Optimaliza\u010Dn\u00ED proces prob\u00EDh\u00E1 ve spojit\u00E9m \u010Dase s vyu\u017Eit\u00EDm prom\u011Bnn\u00E9 velikosti populace. Motivac\u00ED pro pou\u017E\u00EDv\u00E1n\u00ED a vylep\u0161ov\u00E1n\u00ED t\u011Bchto metod optimalizace a tvorby model\u016F je z\u00EDsk\u00E1n\u00ED lep\u0161\u00EDch, kvalitn\u011Bj\u0161\u00EDch model\u016F pro popis dat nebo syst\u00E9m\u016F a z\u00EDsk\u00E1v\u00E1n\u00ED nov\u00FDch znalost\u00ED z t\u011Bchto model\u016F."@cs . "Nature Inspired Methods for Models Creation and Optimization"@en . . . .