| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

Show document Help

Title:Primerjalna analiza hibridno rekurentnega modela in transformerskega modela globokega učenja za napovedovanje delniškega trga
Authors:ID Strahan, Marcel (Author)
ID Šumak, Boštjan (Mentor) More about this mentor... New window
Files:.pdf VS_Strahan_Marcel_2026.pdf (1,99 MB)
MD5: 60D6D6F0931DBF5DF23F9E61685DE684
 
Language:Slovenian
Work type:Bachelor thesis/paper
Typology:2.11 - Undergraduate Thesis
Organization:FERI - Faculty of Electrical Engineering and Computer Science
Abstract:Napovedovanje gibanja cen je izziv zaradi nelinearnosti in nestanovitnosti finančnih trgov. Cilj zaključnega dela je primerjati hibridni model LSTM-GRU z modelom Transformer za napovedovanje zaključnih cen delnice »AMZN« na podlagi podatkov v obdobju 2021–2025. Oba modela ovrednotimo v izhodiščni konfiguraciji in po Bayesovi optimizaciji z regresijskimi metrikami in s smerno natančnostjo na testni množici 221 napovednih vzorcev ter v različnih tržnih režimih. Optimizirani model LSTM-GRU je dosegel najnižji RMSE 4,40 USD in MAE 3,20 USD med vsemi konfiguracijami. Smerna natančnost vseh konfiguracij se giblje okoli 50 % in ni statistično značilna, zato rezultati ne dokazujejo zanesljivega napovedovanja smeri gibanja cene.
Keywords:Globoko učenje, LSTM-GRU, Transformer, napovedovanje delniških cen, Bayesova optimizacija
Place of publishing:Maribor
Year of publishing:2026
PID:20.500.12556/DKUM-99581 New window
Publication date in DKUM:24.09.2026
Views:161
Downloads:6
Metadata:XML DC-XML DC-RDF
Categories:KTFMB - FERI
:
Copy citation
  
Average score:(0 votes)
Your score:Voting is allowed only for logged in users.
Share:Bookmark and Share



Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Link:http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:20.08.2026

Secondary language

Language:English
Title:Comparative analysis of hybrid recurrent and transformer deep learning models for stock market forecasting
Abstract:Predicting price movements is challenging due to the nonlinearity and high volatility of financial markets. This thesis compares the hybrid LSTM-GRU model with the Transformer model for predicting the closing price of the »AMZN« stock using data from 2021 to 2025. Both models are evaluated in their baseline configurations and after Bayesian optimization using regression metrics and directional accuracy on a test set of 221 forecasting samples and across different market regimes. The optimized LSTM-GRU achieved the lowest RMSE of USD 4.40 and MAE of USD 3.20 among all configurations. The directional accuracy for all configurations remained around 50 % and was not statistically significant, indicating that the results do not demonstrate reliable forecasting of the direction of price movements.
Keywords:Deep learning, LSTM-GRU, Transformer, stock price forecasting, Bayesov optimization


Comments

Leave comment

You must log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back
Logos of partners University of Maribor University of Ljubljana University of Primorska University of Nova Gorica