PORTAL: Scalable Tabular Foundation Models via Content-Specific Tokenization

Autor:	Spinaci, Marco, Polewczyk, Marek, Hoffart, Johannes, Kohler, Markus C., Thelin, Sam, Klein, Tassilo
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning
Druh dokumentu:	Working Paper
Popis:	Self-supervised learning on tabular data seeks to apply advances from natural language and image domains to the diverse domain of tables. However, current techniques often struggle with integrating multi-domain data and require data cleaning or specific structural requirements, limiting the scalability of pre-training datasets. We introduce PORTAL (Pretraining One-Row-at-a-Time for All tabLes), a framework that handles various data modalities without the need for cleaning or preprocessing. This simple yet powerful approach can be effectively pre-trained on online-collected datasets and fine-tuned to match state-of-the-art methods on complex classification and regression tasks. This work offers a practical advancement in self-supervised learning for large-scale tabular data. Comment: Accepted at Table Representation Learning Workshop at NeurIPS 2024
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2410.13516 Zobrazit plný text záznamu View this record from Arxiv