A tabular data generation framework guided by downstream tasks optimization
Abstract Recently, generative models have been gradually emerging into the extended dataset field, showcasing their advantages. However, when it comes to generating tabular data, these models often fail to satisfy the constraints of numerical columns, which cannot generate high-quality datasets that...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-07-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-024-65777-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832585744013590528 |
|---|---|
| author | Fengwei Jia Hongli Zhu Fengyuan Jia Xinyue Ren Siqi Chen Hongming Tan Wai Kin Victor Chan |
| author_facet | Fengwei Jia Hongli Zhu Fengyuan Jia Xinyue Ren Siqi Chen Hongming Tan Wai Kin Victor Chan |
| author_sort | Fengwei Jia |
| collection | DOAJ |
| description | Abstract Recently, generative models have been gradually emerging into the extended dataset field, showcasing their advantages. However, when it comes to generating tabular data, these models often fail to satisfy the constraints of numerical columns, which cannot generate high-quality datasets that accurately represent real-world data and are suitable for the intended downstream applications. Responding to the challenge, we propose a tabular data generation framework guided by downstream task optimization (TDGGD). It incorporates three indicators into each time step of diffusion generation, using gradient optimization to align the generated fake data. Unlike the traditional strategy of separating the downstream task model from the upstream data synthesis model, TDGGD ensures that the generated data has highly focused columns feasibility in upstream real tabular data. For downstream task, TDGGD strikes the utility of tabular data over solely pursuing statistical fidelity. Through extensive experiments conducted on real-world tables with explicit column constraints and tables without explicit column constraints, we have demonstrated that TDGGD ensures increasing data volume while enhancing prediction accuracy. To the best of our knowledge, this is the first instance of deploying downstream information into a diffusion model framework. |
| format | Article |
| id | doaj-art-de8307375287441bb552987a10530cb0 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-de8307375287441bb552987a10530cb02025-01-26T12:35:16ZengNature PortfolioScientific Reports2045-23222024-07-0114111410.1038/s41598-024-65777-9A tabular data generation framework guided by downstream tasks optimizationFengwei Jia0Hongli Zhu1Fengyuan Jia2Xinyue Ren3Siqi Chen4Hongming Tan5Wai Kin Victor Chan6Tsinghua Shenzhen International Graduate School, Tsinghua UniversityTsinghua Shenzhen International Graduate School, Tsinghua UniversitySchool of Mechanical Engineering, Anhui University of TechnologyTsinghua Shenzhen International Graduate School, Tsinghua UniversityTsinghua Shenzhen International Graduate School, Tsinghua UniversityTsinghua Shenzhen International Graduate School, Tsinghua UniversityTsinghua Shenzhen International Graduate School, Tsinghua UniversityAbstract Recently, generative models have been gradually emerging into the extended dataset field, showcasing their advantages. However, when it comes to generating tabular data, these models often fail to satisfy the constraints of numerical columns, which cannot generate high-quality datasets that accurately represent real-world data and are suitable for the intended downstream applications. Responding to the challenge, we propose a tabular data generation framework guided by downstream task optimization (TDGGD). It incorporates three indicators into each time step of diffusion generation, using gradient optimization to align the generated fake data. Unlike the traditional strategy of separating the downstream task model from the upstream data synthesis model, TDGGD ensures that the generated data has highly focused columns feasibility in upstream real tabular data. For downstream task, TDGGD strikes the utility of tabular data over solely pursuing statistical fidelity. Through extensive experiments conducted on real-world tables with explicit column constraints and tables without explicit column constraints, we have demonstrated that TDGGD ensures increasing data volume while enhancing prediction accuracy. To the best of our knowledge, this is the first instance of deploying downstream information into a diffusion model framework.https://doi.org/10.1038/s41598-024-65777-9 |
| spellingShingle | Fengwei Jia Hongli Zhu Fengyuan Jia Xinyue Ren Siqi Chen Hongming Tan Wai Kin Victor Chan A tabular data generation framework guided by downstream tasks optimization Scientific Reports |
| title | A tabular data generation framework guided by downstream tasks optimization |
| title_full | A tabular data generation framework guided by downstream tasks optimization |
| title_fullStr | A tabular data generation framework guided by downstream tasks optimization |
| title_full_unstemmed | A tabular data generation framework guided by downstream tasks optimization |
| title_short | A tabular data generation framework guided by downstream tasks optimization |
| title_sort | tabular data generation framework guided by downstream tasks optimization |
| url | https://doi.org/10.1038/s41598-024-65777-9 |
| work_keys_str_mv | AT fengweijia atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT honglizhu atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT fengyuanjia atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT xinyueren atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT siqichen atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT hongmingtan atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT waikinvictorchan atabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT fengweijia tabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT honglizhu tabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT fengyuanjia tabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT xinyueren tabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT siqichen tabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT hongmingtan tabulardatagenerationframeworkguidedbydownstreamtasksoptimization AT waikinvictorchan tabulardatagenerationframeworkguidedbydownstreamtasksoptimization |