Naslov
Stock analysis and forecast using neural networks (INA-R-A)

Autori
Cingula, Domagoj ; Primorac, Dinko ; Borović, Franjo

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
1st International and Scientific Conference "Economic and Social Development" : proceedings / Filipović, David ; Goltnik Urnaut, Anita - Celje : Faculty of Commercial and Business Sciences Celje, 2012, 399-408

ISBN
978-961-6825-49-8

Skup
International and Scientific Conference (1 ; 2012)

Mjesto i datum
Frankfurt na Majni, Njemačka, 12.04.2012. - 13.04.2012

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Analysis ; Data mining ; Forecast ; JMP ; Neural Networks

Sažetak
Certain patterns are visible in every capital market, certain features characterize every capital market. Looking at both of the above plus if usual impacts are included some movements can be predicted. Past trends, statistics and probability or stabile and strong favorites are not a guarantee of future trends, yet all mentioned represent strong support in process of making forecast. Neural networks, that simulate the brain cells, represent quality tool for such a prediction and analysis. Statistical computer software JMP allows easy predictions based on extremely large databases. INA Industrija nafte d.d. is Croatian oil company and one of the main business entities in the country. INA-R-A stands for INA Industrija nafte d.d. stock symbol on Zagreb Stock Exchange (ZSE) where their stocks are listed. Data used in the analysis and as a base for making forecast include period between December 2006. and February 2011. Trading with INA-R-A was set on hold from 28th of April 2011 to 19th of December 2011 by Croatian Financial Services Supervisory Agency (HANFA ) and that is the reason why analysis and forecast doesn't include up to date information. Analysis and forecast base is made of all trading data from the mentioned period (Open , Close or Average, High and Low price plus Turnover) with CROBEX (ZSE main index), Oil prices and Baltic Dry Index history included. Method of using Neural Networks and software used in the paper (JMP) is easily applied to any other field, scientific or other, where there is a need of spotting the connection between different variables in large data bases.

Izvorni jezik
Engleski

Znanstvena područja
Ekonomija

POVEZANOST RADA