Pregled bibliografske jedinice broj: 74620
Saturation filtering for noise and outlier detection
Saturation filtering for noise and outlier detection // Active Learning - Database Sampling - Experimental Design: Views on Instance Selection / Scheffer, Tobias; Wrobel, Stefan (ur.).
Freiburg: ECKM/PKDD Conference, 2001. str. 21-24 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 74620 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Saturation filtering for noise and outlier detection
Autori
Lavrač, Nada ; Gamberger, Dragan
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Active Learning - Database Sampling - Experimental Design: Views on Instance Selection
/ Scheffer, Tobias; Wrobel, Stefan - Freiburg : ECKM/PKDD Conference, 2001, 21-24
Skup
Workshop at European Conference on Machine Learning
Mjesto i datum
Freiburg, Njemačka, 05.09.2001
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
noise detection; saturation filtering
Sažetak
Instance selection is an important part of KDD process. It is aimed at finding a representative data subset that can replace the original dataset, still solving a data mining task as if the whole dataset were used. For most data mining task, including classification tasks, the selected dataset should preferably exclude noisy instances, aimed at improving the predictive accuracy and applicability of induced knowledge significantly depend on appropriate noise handling procedures.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika
POVEZANOST RADA
