Umjetni neuronski sustav odlučivanja o kupovini sirove nafte

doctoral thesis

• Title (english): Artificial neural decision making system for purchase of crude oil
• Author: Predrag Domijan
• Mentor: Damir Kalpić (mentor)
• Committee member: Vedran Mornar (član povjerenstva)
• Committee member: Damir Kalpić (član povjerenstva)
• Committee member: Vesna Dušak (član povjerenstva)
• Committee member: Ivan Petrović (član povjerenstva)
• Committee member: Bojana Dalbelo-Bašić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Applied Mathematics); Zagreb
• Defense date and country: 2004-11-11, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing; Process Computing
• Universal decimal classification (UDC): 004 - Computer science and technology. Computing. Data processing

Abstract

U disertaciji je razmatrana špekulativna aktivnost kupovine sirove nafte za rafineriju na spot tržištu. Cilj je ostvariti dobit udobnosti posjedovanja sirovine, te špekulativnu dobit kao smanjenje cijene sirovine. Projektirana je umjetna neuronska mreža kao sastavni dio sustava podrške donošenju odluka. S obzirom na teško, zapravo nemoguće, predviđanje budućih spot cijena, inovativni pristup je primijenjen u modeliranju umjetne neuronske mreže. Kao izlazna varijabla neuronske mreže ne koristi se buduća spot cijena sirove nafte već odluka koja predstavlja malu, srednju ili veliku količinu sirove nafte koju se kupuje. Neuronska mreža se trenira i izračunava odluke u periodu od osam godina (1996.-2003.). Novi podaci se uzimaju dnevno, a umjetna neuronska mreža se trenira svaka tri mjeseca. Ulazni vektor za umjetnu neuronsku mrežu heuristički je izveden isključivo iz historijskih podataka - spot cijena sirove nafte. Dobivena je prosječna dobit od 1.6 posto za cijeli period (1996.-2003.). Projektirana metodologija primjenjiva je na aplikacijama gdje se energija ili sirovina kupuje na spot tržištu u cilju potrošnje u narednom kratkoročnom periodu do tri mjeseca.

Abstract (english)

The dissertation examines buying of crude oil for a refinery on spot market based on a speculative trading strategy. The goal is to increase convenience yield and speculative gain by reducing the cost of purchased crude oil. An Artificial Neural Network (ANN) was designed as a part of Decision Support System (DSS). Because of difficulties in forecasting the crude oil spot price, an innovative approach was applied in modeling the ANN. ANN output variables are not future spot prices, but decisions that represent small, mid or large quantity of crude oil to be bought. The ANN was trained and tested on decisions during the eight years time period (19962003). The crude oil spot prices are collected daily. The model is trained every three months. Input variables to the ANN were heuristically derived from the historical spot prices of crude oil. About 1.6 % speculative profits have been obtained for the time period (1996-2003). Designed methodology presented in this work could be implemented on any application where the energy or raw material is being bought on spot market in order to be stored for consumption within a short time period of up to three months.

• Keywords: Špekulativna strategija trgovanja; sirova nafta; cijene nafte; umjetna neuronska mreža; sustav podrške donošenju odluke; rafinerija
• Keywords (english): Speculative trading strategy; crude oil; oil prices; artificial neural network; decision support system; refinery
• Language: croatian
• URN:NBN: urn:nbn:hr:168:946193
• Study programme: Title: Computer Science; Study programme type: university; Study level: postgraduate; Academic / professional title: Doktor znanosti (Doktor znanosti)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-06-12 12:58:07