Decision-Aware Conditional GANs for Time Series Data
We introduce the decision-aware time-series conditional generative adversarial network (DAT-CGAN), a method for the generation of time-series data that is designed to support decision-making. The framework adopts a multi-Wasserstein loss on decision-related quantities and an overlapped block-samplin...
Main Authors: | , , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
ACM|4th ACM International Conference on AI in Finance
2023
|
Online Access: | https://hdl.handle.net/1721.1/153133 |
_version_ | 1811081397899624448 |
---|---|
author | Sun, He Deng, Zhun Chen, Hui Parkes, David |
author2 | Sloan School of Management |
author_facet | Sloan School of Management Sun, He Deng, Zhun Chen, Hui Parkes, David |
author_sort | Sun, He |
collection | MIT |
description | We introduce the decision-aware time-series conditional generative adversarial network (DAT-CGAN), a method for the generation of time-series data that is designed to support decision-making. The framework adopts a multi-Wasserstein loss on decision-related quantities and an overlapped block-sampling approach for sample
efficiency. We characterize the generalization properties of DAT-CGAN and in application to a multi-period portfolio choice problem and financial time series data, we demonstrate better training stability and generative quality in regard to both raw data and decision-related quantities than strong GAN-based baselines. |
first_indexed | 2024-09-23T11:46:03Z |
format | Article |
id | mit-1721.1/153133 |
institution | Massachusetts Institute of Technology |
language | English |
last_indexed | 2024-09-23T11:46:03Z |
publishDate | 2023 |
publisher | ACM|4th ACM International Conference on AI in Finance |
record_format | dspace |
spelling | mit-1721.1/1531332024-01-11T21:04:22Z Decision-Aware Conditional GANs for Time Series Data Sun, He Deng, Zhun Chen, Hui Parkes, David Sloan School of Management We introduce the decision-aware time-series conditional generative adversarial network (DAT-CGAN), a method for the generation of time-series data that is designed to support decision-making. The framework adopts a multi-Wasserstein loss on decision-related quantities and an overlapped block-sampling approach for sample efficiency. We characterize the generalization properties of DAT-CGAN and in application to a multi-period portfolio choice problem and financial time series data, we demonstrate better training stability and generative quality in regard to both raw data and decision-related quantities than strong GAN-based baselines. 2023-12-11T21:33:51Z 2023-12-11T21:33:51Z 2023-11-27 2023-12-01T08:47:38Z Article http://purl.org/eprint/type/ConferencePaper 979-8-4007-0240-2 https://hdl.handle.net/1721.1/153133 Sun, He, Deng, Zhun, Chen, Hui and Parkes, David. 2023. "Decision-Aware Conditional GANs for Time Series Data." PUBLISHER_POLICY en https://doi.org/10.1145/3604237.3626855 Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. The author(s) application/pdf ACM|4th ACM International Conference on AI in Finance Association for Computing Machinery |
spellingShingle | Sun, He Deng, Zhun Chen, Hui Parkes, David Decision-Aware Conditional GANs for Time Series Data |
title | Decision-Aware Conditional GANs for Time Series Data |
title_full | Decision-Aware Conditional GANs for Time Series Data |
title_fullStr | Decision-Aware Conditional GANs for Time Series Data |
title_full_unstemmed | Decision-Aware Conditional GANs for Time Series Data |
title_short | Decision-Aware Conditional GANs for Time Series Data |
title_sort | decision aware conditional gans for time series data |
url | https://hdl.handle.net/1721.1/153133 |
work_keys_str_mv | AT sunhe decisionawareconditionalgansfortimeseriesdata AT dengzhun decisionawareconditionalgansfortimeseriesdata AT chenhui decisionawareconditionalgansfortimeseriesdata AT parkesdavid decisionawareconditionalgansfortimeseriesdata |