TabKANet: Tabular Data Modeling with Kolmogorov-Arnold Network and Transformer
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910628028874752 |
|---|---|
| author | Gao, Weihao Gong, Zheng Deng, Zhuo Rong, Fuju Chen, Chucheng Ma, Lan |
| author_facet | Gao, Weihao Gong, Zheng Deng, Zhuo Rong, Fuju Chen, Chucheng Ma, Lan |
| contents | Tabular data is the most common type of data in real-life scenarios. In this study, we propose the TabKANet model for tabular data modeling, which targets the bottlenecks in learning from numerical content. We constructed a Kolmogorov-Arnold Network (KAN) based Numerical Embedding Module and unified numerical and categorical features encoding within a Transformer architecture. TabKANet has demonstrated stable and significantly superior performance compared to Neural Networks (NNs) across multiple public datasets in binary classification, multi-class classification, and regression tasks. Its performance is comparable to or surpasses that of Gradient Boosted Decision Tree models (GBDTs). Our code is publicly available on GitHub: https://github.com/AI-thpremed/TabKANet. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_08806 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | TabKANet: Tabular Data Modeling with Kolmogorov-Arnold Network and Transformer Gao, Weihao Gong, Zheng Deng, Zhuo Rong, Fuju Chen, Chucheng Ma, Lan Machine Learning Artificial Intelligence Tabular data is the most common type of data in real-life scenarios. In this study, we propose the TabKANet model for tabular data modeling, which targets the bottlenecks in learning from numerical content. We constructed a Kolmogorov-Arnold Network (KAN) based Numerical Embedding Module and unified numerical and categorical features encoding within a Transformer architecture. TabKANet has demonstrated stable and significantly superior performance compared to Neural Networks (NNs) across multiple public datasets in binary classification, multi-class classification, and regression tasks. Its performance is comparable to or surpasses that of Gradient Boosted Decision Tree models (GBDTs). Our code is publicly available on GitHub: https://github.com/AI-thpremed/TabKANet. |
| title | TabKANet: Tabular Data Modeling with Kolmogorov-Arnold Network and Transformer |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2409.08806 |