Institutional Repository of Vilnius University
Sintetinių finansinių duomenų naudojimas automatizuotoje prekyboje: Prekybos įrankio kūrimas /
Abstract
dc:descriptionThis thesis investigates the potential of using synthetic financial time-series data generated via machine learning in algorithmic trading backtesting models. The synthetic financial time-series data itself was generated by one of the extensions of the generative adversarial network architecture. Specifically, the conditional time-series generative adversarial network architecture with long short-term memory networks was used. It was chosen due to its ability to learn the temporal tendencies of the real financial time-series data, while also being customizable due to the conditional inputs. The similarity between the synthetic and real financial time-series was shown by using the known statistical properties of financial instruments such as loss-gain asymmetry and volatility clustering. Using the mean-reversion algorithmic method in tandem with metrics such as Bollinger bands, Z-Score and the relative strength index, it was shown that indeed synthetic financial time-series data is a potentially viable substitute in backtesting models. Also a special tool was developed to load the models.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Šamšonas, Matas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD239032424&prefLang=en_US