Sustav za predviđanje kretanja vrijednosti kriptovaluta temeljen na dubokom učenju

master's thesis

• Title (english): System for Prediction of Criptocurency Value Changes Based on Deep Learning
• Author: Zvonimir Vujnović
• Mentor: Tomislav Hrkać (mentor)
• Committee member: Tomislav Hrkać (predsjednik povjerenstva)
• Committee member: Zoran Kalafatić (član povjerenstva)
• Committee member: Jan Šnajder (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-02, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom radu opisane su kriptovalute i povratne neuronske mreže. Kriptovalute su "digitalne" valute. Opisano je što su i kako su izvedene te koja je njihova svrha. Nakon toga opisano je što su povratne neuronske mreže, problemi koje rješavaju, arhitekture koje postoje i njihove varijacije. U zadnjem poglavlju opisani su modeli pomoću kojih se predviđa cijena kriptovaluta. Model se sastoji od LSTM ćelija. To su posebne vrste RNN ćelija koje imaju bolje dugoročno pamćenje od same RNN ćelije. Iako je model vrlo jednostavan, dobro uči i interpretira podatke, ali nije dovoljno pouzdan za davati u potpunosti točne predikcije jer je tržište kriptovaluta jako volatilno.

Abstract (english)

This paper describes cryptocurrencies and recurrent neural networks. Cryptocurrencies are “digital” currencies. It describes what they are, how they were made and what their purpose is. It is then described what recurrent neural networks are, the problems they solve, the architectures that exist, and their variations. The last chapter describes the models used to predict the price of cryptocurrencies. The model consists of LSTM cells. These are special types of RNN cells that have better long-term memory than the RNN cell itself. Although the model is very simple, it learns and interprets the data well, but it is not reliable enough to give completely accurate predictions because the cryptocurrency market is very volatile.

• Keywords: Kriptovalute; predviđanje; povratne neuronske mreže
• Keywords (english): Cryptocurrencies; prediction; reccurent neural networks
• Language: croatian
• URN:NBN: urn:nbn:hr:168:636298
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 16:31:01