Financial time series augmentation using transformer based GAN architecture

Andrzej Podobiński 1 , Jarosław A. Chudziak 1 1 Faculty of Electronics and Information Technology, Warsaw University of Technology, Warsaw, Poland [email protected], [email protected] · 2026

Read Financial time series augmentation using transformer based GAN architecture by Andrzej Podobiński 1 , Jarosław A. Chudziak 1 1 Faculty of Electronics and Information Technology, Warsaw University of Technology, Warsaw, Poland [email protected], [email protected] (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-17865.txt.

Description

Andrzej Podobiński and Jarosław Chudziak show how generative rival networks can boost scarce financial time series data for forecasting. Forecasting matters across fields from engineering to economics, yet jumpy finance data shortage yields suboptimal training and poor carry-over. The core challenge is reliably boosting scarce financial time series to lift deep learning forecast accuracy.

The authors train a long short-term memory forecast model on data boosted with fake data from a transformer-based GAN, TTS-GAN, and find clearly improved forecast accuracy against real data alone. Results hold across Bitcoin and S&P500 price data and varied forecast spans. A novel, time series targeted quality metric combining dynamic time warping with a tweaked deep dataset gap measure reliably monitors training progress and judges generated data quality. The findings support GAN-based boosting as a practical tool for financial prediction.

Catalog record

CREATORAndrzej Podobiński 1 , Jarosław A. Chudziak 1 1 Faculty of Electronics and Information Technology, Warsaw University of Technology, Warsaw, Poland [email protected], [email protected]
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-17865

Browse the full library · About the Tower of Babel