DocumentCode :
2134976
Title :
Why fuzzy in data mining?
Author :
Vámos, Tibor
Author_Institution :
Comput. & Autom. Res. Inst., Hungarian Acad. of Sci.
fYear :
2003
fDate :
24-24 Sept. 2003
Firstpage :
46
Lastpage :
49
Abstract :
Fuzzy - as has been emphasized by Zadeh - is a human interface, a link between human judgement and machine information processing. It represents an approach for experimentation with the sensitivity of human classification, i.e. the notion of adding tacit semantic knowledge (Polanyi) to search procedures in data banks. A brute force algorithm operates on data grading and creating clusters, class structures based on highly variable numbers and the setting of different functions for the refinement of class granulation (sensitivity classes, e.g. Weber-Fechner perception law), tolerances etc. The method was successfully applied in a medical database for the research of hidden interactions in brain development and in investigating emerging phenomena of European countries joining currently the EU
Keywords :
data mining; fuzzy systems; search problems; brute force algorithm; data banks; data mining; fuzzy system; human judgement; machine information processing; medical database; search procedures; semantic knowledge; sensitivity class granulation; Automation; Blood pressure; Clustering algorithms; Computer interfaces; Data mining; Databases; Humans; Information processing; Mirrors; Psychology;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Uncertainty Modeling and Analysis, 2003. ISUMA 2003. Fourth International Symposium on
Conference_Location :
College Park, MD
Print_ISBN :
0-7695-1997-0
Type :
conf
DOI :
10.1109/ISUMA.2003.1236139
Filename :
1236139
Link To Document :
بازگشت