Original title: Image classification - decision design
Translated title: Klasifikace obrazu - návrh klasifikátoru
Authors: Klimešová, Dana ; Ocelíková, E.
Document type: Papers
Conference/Event: Process Control 2004 - ŘÍP 2004 /6./, Kouty nad Desnou (CZ), 2004-06-08 / 2004-06-11
Year: 2004
Language: eng
Abstract: [eng] [cze]
This paper presents overview of the mostly used classification approaches and shows the possibility of the approach based on the maximum entropy theory that can optimally describe the spatial data distribution and gives actual error estimation. The approach is highly recommended in case we are not able to test the assumptions of given method. Článek prezentuje nejčastěji užívané klasifikační metody a ukazuje na možné využití postupů založených na teorii maximální entropie v případech, kdy nemáme možnost ověřit předpoklady dané metody a zaručit odhad chyby dané metody
Keywords: distribution function; error distribution function; spatial data classification
Project no.: CEZ:AV0Z1075907 (CEP)
Host item entry: Proceedings of the 6th International Scientific -Technical Conference. Process Control 2004, ISBN 80-7194-662-1

Institution: Institute of Information Theory and Automation AS ČR (web)
Document availability information: Fulltext is available at the institute of the Academy of Sciences.
Original record: http://hdl.handle.net/11104/0013466

Permalink: http://www.nusl.cz/ntk/nusl-20068


The record appears in these collections:
Research > Institutes ASCR > Institute of Information Theory and Automation
Conference materials > Papers