Pregled bibliografske jedinice broj: 854413
Optimal number of clusters provided by k-means and E-M algorithm
Optimal number of clusters provided by k-means and E-M algorithm // Proceedings of 8th International Scientific and Expert Conference of the International TEAM Society / AlumniPress (ur.).
Trnava: Faculty of Materials Science and Technology in Trnava, 2016. str. 286-291 (ostalo, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 854413 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Optimal number of clusters provided by k-means and E-M algorithm
Autori
Novoselac, Vedran ; Pavić, Zlatko
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings of 8th International Scientific and Expert Conference of the International TEAM Society
/ AlumniPress - Trnava : Faculty of Materials Science and Technology in Trnava, 2016, 286-291
ISBN
9788080962371
Skup
8th International Scientific and Expert Conference of the International TEAM Society
Mjesto i datum
Trnava, Slovačka, 19.10.2016. - 21.10.2016
Vrsta sudjelovanja
Ostalo
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
k-means ; expectation maximization ; grouping index
Sažetak
The paper considers the problem of determining the optimal number of clusters in data set by grouping index. The problem of clustering are provided with k-means and E-M (Expectation Maximization) algorithm. In addition to well- known indexes that are frequently used, two new indexes are presented. New indexes are based on the orthogonal distances from data to the line which represent corresponding cluster in the partition obtained with mentioned algorithms.
Izvorni jezik
Engleski
Znanstvena područja
Matematika, Elektrotehnika
POVEZANOST RADA
Ustanove:
Strojarski fakultet, Slavonski Brod,
Sveučilište J. J. Strossmayera u Osijeku