20.500.12556/DKUM-81480 Uporaba umetne inteligence pri upravljanju portfelja delnic Artificial intelligence for managing the portfolio of stocks Izziv dela predstavlja snovanje, načrtovanje in praktična izvedba avtomatiziranega trgovalnega sistema, ki neodvisno in brez posredovanja uporabnikov sprejema in izvaja trgovalne odločitve. Jedro trgovalnega sistema predstavlja trgovalna strategija, ki spremlja pretekle ter aktualne podatke borznih kotacij, izvaja tehnično analizo in, če je tega sposobna, se prilagaja sprotnim razmeram na finančnih trgih. Obravnavamo dve skupini trgovalnih strategij, klasične, ki niso sposobne sprotnega prilagajanja niti učenja, in dve trgovalni strategiji na osnovi naprednih algoritmov umetne inteligence, eno izmed njih predstavnico umetnih nevronskih mrež najnovejše tretje generacije. Izvedemo obširna simulacijska eksperimentiranja na osnovi nemškega delniškega trga v zadnjih desetih letih, zasnujemo in izvedemo pa tudi eksperimentiranja na namenski strojni opremi, ki močno pohitri kompleksnost časovnega izvajanja, ter eksperimentiranja na analognem elektronskem vezju, s pomočjo katerega se podrobno seznanimo z načinom propagiranja informacij umetnih nevronskih mrež tretje generacije. Rezultati eksperimentov prinašajo tako vsebinske kot tehnične ugotovitve, najpomembnejšo med njimi, da se enoten model ki hkrati trguje z večjim številom finančnih instrumentov obnaša podobno kot kopica posamično prilagojenih modelov na točno določen finančni instrument, kakor tudi novo ugotovljene izkušnje vezane na propagiranje in izrabo najnovejše generacije umetnih nevronskih mrež. Challenge of this work is about the design, planning and practical implementation of the automated trading system that independently and user-interference free generates the trading decisions and realizes them on the open market. The core of the trading system is a trading strategy that monitors past and current financial data, performs the technical analyses and, if capable adapts to the latest conditions on financial markets. Two groups of the trading strategies, the classics, traditionally not capable of adapting nor learning, and two adapting and learning capable trading strategies, based on the latest artificial intelligence methods, one of them a representative of the third-generation neural networks, are implemented. Comprehensive simulation experiments and tests are concluded using the data on German stock market in past ten years, with additional digital and purely analogue hardware experiments on the dedicated equipment, that demand significantly lower time complexities on one hand, and offer an outstanding chance to get familiarized with the concept of information propagation in the third-generation artificial neural networks. Results of experiments communicate both substantive and technical findings, more important among them, that the universal model that manages several financial instruments concurrently behaves similar as a bunch of specific models that are specialized for only a single financial instrument at a time, as well as newly discovered experiences on propagation and exploitation of the latest third-generation neural networks. umetna inteligenca portfelj delnic umetne nevronske mreže mehanski trgovalni sistem artificial intelligence portfolio of stocks artificial neural networks mechanical trading system true true false [D. Fister] Slovenski jezik Angleški jezik Doktorsko delo/naloga 2022-03-31 09:40:00 2022-11-14 10:56:08 2022-11-15 03:14:11 0000-00-00 00:00:00 2022 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 336.76:004.8(043.3) 129178883 DOK_Fister_Dusan_2022.pdf DOK_Fister_Dusan_2022.pdf 1 7557B3FCDC89CCE2587FE24D4DF296EB 833ad113cdb78a6165aedaa1846fcc89f328d7a8fba1866ae927a92f818dd5f8 e926387d-3dad-11ed-9af8-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=163445 Ekonomsko-poslovna fakulteta 0 0 0