AAAI 2025technical0 citations
Transfer Learning in Financial Time Series with Gramian Angular Field (Student Abstract)
Hou-Wan Long, On-In Ho, Qi-Qiao He, Yain-Whar Si
Abstract
Transfer learning enhances model performance in financial time series by leveraging data from related domains. The selection of appropriate source domains is crucial to avoid negative transfer. We propose using Gramian Angular Field (GAF) transformations to improve time series similarity functions for better domain alignment. Extensive experiments with DNN and LSTM models show that GAF-based similarity functions, specifically Coral (GAF) for DNN and CMD (GAF) for LSTM, significantly reduce prediction errors, demonstrating their effectiveness in complex financial environments.
BibTeX
@article{Long_Ho_He_Si_2025, title={Transfer Learning in Financial Time Series with Gramian Angular Field (Student Abstract)}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/35272}, DOI={10.1609/aaai.v39i28.35272}, abstractNote={Transfer learning enhances model performance in financial time series by leveraging data from related domains. The selection of appropriate source domains is crucial to avoid negative transfer. We propose using Gramian Angular Field (GAF) transformations to improve time series similarity functions for better domain alignment. Extensive experiments with DNN and LSTM models show that GAF-based similarity functions, specifically Coral (GAF) for DNN and CMD (GAF) for LSTM, significantly reduce prediction errors, demonstrating their effectiveness in complex financial environments.}, number={28}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Long, Hou-Wan and Ho, On-In and He, Qi-Qiao and Si, Yain-Whar}, year={2025}, month={Apr.}, pages={29418-29420} }