"3" . . "Vyu\u017Eit\u00ED modelu neuronov\u00E9 s\u00EDt\u011B v modelov\u00E1n\u00ED pr\u016Ftok\u016F" . "VTEI" . "50" . . "1"^^ . "RIV/00020711:_____/08:00002136" . . "Application of neural network model in discharge modelling"@en . . . . "Application of neural network model in discharge modelling"@en . "[757651431F38]" . . . . . "Rainfall-runoff process in hydrological system can be successfully modelled by black-box models. The representative black-box model is an Artificial Neural Network model (ANN) that was used in interdisciplinary problem treatment in the last decades. Applying ANN models is necessary to choose properly the input variables and include only those that correlate with output and can successfully explain the output. Engaging the irrelevant or redundant input variables can lead to too complex structures of ANN model and its overparametrization. Instead of measured rainfalls, a rainfall data modified to Antecedent Precipitation Indexes (API) can be used as input variables. The aim of this article is the evaluation of optimal number of inputs into the ANN rainfall- runoff model."@en . "0322-8916" . "P\u0159i modelov\u00E1n\u00ED sr\u00E1\u017Eko-odtokov\u00E9ho procesu v hydrologick\u00E9m syst\u00E9mu lze \u00FAsp\u011B\u0161n\u011B vyu\u017E\u00EDvat black-boxov\u00E9 modely. Typick\u00FDm z\u00E1stupcem t\u011Bchto model\u016F je model um\u011Bl\u00E9 neuronov\u00E9 s\u00EDt\u011B (ANN - Artificial Neural Network), kter\u00FD byl v posledn\u00EDch desetilet\u00EDch vyu\u017Eit p\u0159i \u0159e\u0161en\u00ED probl\u00E9m\u016F v r\u016Fzn\u00FDch v\u011Bdn\u00EDch oborech. P\u0159i aplikaci model\u016F um\u011Bl\u00FDch neuronov\u00FDch s\u00EDt\u00ED je nutn\u00E9 dob\u0159e uv\u00E1\u017Eit veli\u010Diny, kter\u00E9 budou do modelu vstupovat, a za\u0159adit jen ty, kter\u00E9 koreluj\u00ED s v\u00FDstupem a maj\u00ED tedy potenci\u00E1l \u00FAsp\u011B\u0161n\u011B jej vysv\u011Btlit. Za\u0159azen\u00ED nerelevantn\u00EDch nebo nadbyte\u010Dn\u00FDch vstupn\u00EDch veli\u010Din m\u016F\u017Ee v\u00E9st k p\u0159\u00EDli\u0161 komplexn\u00EDm struktur\u00E1m modelu ANN a jeho p\u0159eparametrizov\u00E1n\u00ED. Jako vstupn\u00ED veli\u010Diny mohou b\u00FDt, krom\u011B m\u011B\u0159en\u00FDch sr\u00E1\u017Eek, pou\u017Eita upraven\u00E1 sr\u00E1\u017Ekov\u00E1 data jako jsou ukazatele p\u0159edchoz\u00EDch sr\u00E1\u017Eek (API \u2013 Antecedent precipitation index). Hlavn\u00EDm c\u00EDlem tohoto p\u0159\u00EDsp\u011Bvku je vyhodnocen\u00ED optim\u00E1ln\u00EDho po\u010Dtu vstup\u016F do sr\u00E1\u017Eko-odtokov\u00E9ho modelu ANN." . "2"^^ . . . . "RIV/00020711:_____/08:00002136!RIV08-MZP-00020711" . "P\u0159i modelov\u00E1n\u00ED sr\u00E1\u017Eko-odtokov\u00E9ho procesu v hydrologick\u00E9m syst\u00E9mu lze \u00FAsp\u011B\u0161n\u011B vyu\u017E\u00EDvat black-boxov\u00E9 modely. Typick\u00FDm z\u00E1stupcem t\u011Bchto model\u016F je model um\u011Bl\u00E9 neuronov\u00E9 s\u00EDt\u011B (ANN - Artificial Neural Network), kter\u00FD byl v posledn\u00EDch desetilet\u00EDch vyu\u017Eit p\u0159i \u0159e\u0161en\u00ED probl\u00E9m\u016F v r\u016Fzn\u00FDch v\u011Bdn\u00EDch oborech. P\u0159i aplikaci model\u016F um\u011Bl\u00FDch neuronov\u00FDch s\u00EDt\u00ED je nutn\u00E9 dob\u0159e uv\u00E1\u017Eit veli\u010Diny, kter\u00E9 budou do modelu vstupovat, a za\u0159adit jen ty, kter\u00E9 koreluj\u00ED s v\u00FDstupem a maj\u00ED tedy potenci\u00E1l \u00FAsp\u011B\u0161n\u011B jej vysv\u011Btlit. Za\u0159azen\u00ED nerelevantn\u00EDch nebo nadbyte\u010Dn\u00FDch vstupn\u00EDch veli\u010Din m\u016F\u017Ee v\u00E9st k p\u0159\u00EDli\u0161 komplexn\u00EDm struktur\u00E1m modelu ANN a jeho p\u0159eparametrizov\u00E1n\u00ED. Jako vstupn\u00ED veli\u010Diny mohou b\u00FDt, krom\u011B m\u011B\u0159en\u00FDch sr\u00E1\u017Eek, pou\u017Eita upraven\u00E1 sr\u00E1\u017Ekov\u00E1 data jako jsou ukazatele p\u0159edchoz\u00EDch sr\u00E1\u017Eek (API \u2013 Antecedent precipitation index). Hlavn\u00EDm c\u00EDlem tohoto p\u0159\u00EDsp\u011Bvku je vyhodnocen\u00ED optim\u00E1ln\u00EDho po\u010Dtu vstup\u016F do sr\u00E1\u017Eko-odtokov\u00E9ho modelu ANN."@cs . "Z(MZP0002071101)" . "9;10" . . "Vyu\u017Eit\u00ED modelu neuronov\u00E9 s\u00EDt\u011B v modelov\u00E1n\u00ED pr\u016Ftok\u016F"@cs . . "model input; neural network; small catchment; R software"@en . "Vyu\u017Eit\u00ED modelu neuronov\u00E9 s\u00EDt\u011B v modelov\u00E1n\u00ED pr\u016Ftok\u016F" . "\u0158edinov\u00E1, Jana" . "404703" . "1"^^ . "Vyu\u017Eit\u00ED modelu neuronov\u00E9 s\u00EDt\u011B v modelov\u00E1n\u00ED pr\u016Ftok\u016F"@cs . . "CZ - \u010Cesk\u00E1 republika" .