← Glossary Term
Tabular data
Plain rows-and-columns data, the kind that lives in spreadsheets and databases, and the format AI has been slowest to master.
Tabular data is the least glamorous and most common data in the world: customers in rows, attributes in columns, one value you would like to predict. Credit scores, churn forecasts, medical risk scores and demand planning are all tabular problems.
It is also the odd gap in the current AI story. Language models handle text, images, audio and code well, and lose to decades-old statistical methods here. That gap is why “tabular foundation models” have become a research field of their own, with models such as TabPFN built specifically for tables rather than sentences.
Mentioned in