Pregled bibliografske jedinice broj: 462487
Stock market prediction by combining multiple neural network models for standard technical analysis tools
Stock market prediction by combining multiple neural network models for standard technical analysis tools // 23rd EUROPEAN CONFERENCE ON OPERATIONAL RESEARCH EURO XXIII, BOOK OF ABSTRACTS
Bonn, Njemačka, 2009. str. 90-91 (predavanje, međunarodna recenzija, sažetak, znanstveni)
CROSBI ID: 462487 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Stock market prediction by combining multiple neural network models for standard technical analysis tools
Autori
Kokan, Tonći
Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni
Izvornik
23rd EUROPEAN CONFERENCE ON OPERATIONAL RESEARCH EURO XXIII, BOOK OF ABSTRACTS
/ - , 2009, 90-91
Skup
23rd EUROPEAN CONFERENCE ON OPERATIONAL RESEARCH EURO XXIII
Mjesto i datum
Bonn, Njemačka, 05.07.2009. - 08.07.2009
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
stock market prediction; neural network models; technical analysis techniques
Sažetak
Neural network models are one the most frequently used alternatives to traditional technical analysis tools for stock market prediction. In this paper we combine the two approaches by modeling several widely used technical analysis techniques by neural networks. Similarities and differences between those techniques are then explored by comparing their corresponding neural network models. Algorithms for construction of neural networks as combinations of the obtained models are presented. Performances of the constructed models are tested against performances of the standard techniques.
Izvorni jezik
Engleski
Znanstvena područja
Ekonomija
POVEZANOST RADA
Projekti:
055-0000000-1435 - Matematički modeli u analizi razvoja hrvatskog financijskog tržišta (Aljinović, Zdravka, MZOS ) ( CroRIS)
Ustanove:
Ekonomski fakultet, Split
Profili:
Tonći Kokan
(autor)