Pregled bibliografske jedinice broj: 633391
Supervised feature extraction for tensor objects based on maximization of mutual information
Supervised feature extraction for tensor objects based on maximization of mutual information // Pattern recognition letters, 34 (2013), 13; 1476-1484 doi:10.1016/j.patrec.2013.05.018 (međunarodna recenzija, članak, znanstveni)
CROSBI ID: 633391 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Supervised feature extraction for tensor objects based on maximization of mutual information
Autori
Jukić, Ante ; Filipović, Marko
Izvornik
Pattern recognition letters (0167-8655) 34
(2013), 13;
1476-1484
Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni
Ključne riječi
dimensionality reduction; tensor decomposition; feature extraction; mutual information
Sažetak
Several supervised feature extraction methods for tensor objects have been proposed recently, with applications in recognition of objects, faces and handwritten digits. However, the existing methods usually use only second order statistics of the data, typically through calculation of the within- and between-class scatters. Here we propose a method for supervised feature extraction for tensor objects based on maximization of an approximation of mutual information. In this way we utilize information contained in the higher order statistics of the data. Several experiments show that the proposed method results in highly discriminative features.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika, Računarstvo
POVEZANOST RADA
Projekti:
098-0982903-2558 - Analiza višespektralih podataka (Kopriva, Ivica, MZOS ) ( CroRIS)
Ustanove:
Institut "Ruđer Bošković", Zagreb
Poveznice na cjeloviti tekst rada:
Pristup cjelovitom tekstu rada doi www.sciencedirect.com dx.doi.orgCitiraj ovu publikaciju:
Časopis indeksira:
- Current Contents Connect (CCC)
- Web of Science Core Collection (WoSCC)
- Science Citation Index Expanded (SCI-EXP)
- SCI-EXP, SSCI i/ili A&HCI
- Scopus
