2 ms·
Google's TabFM [1] (and its previous TimesFM) seem the best approach so far on this area 1. https://research.google/blog/introducing-tabfm-a-zero-shot-foundati
by tough 2mo ago
Google's TabFM [1] (and its previous TimesFM) seem the best approach so far on this area
1. https://research.google/blog/introducing-tabfm-a-zero-shot-foundation-model-for-tabular-data/ https://research.google/blog/introducing-tabfm-a-zero-shot-f...
- jeffreyrogers 2mo agoLast time I checked xgboost and lightgbm still outperformed anything NN based for tabular data.
- woofwoofwoof 2mo agoTabPFN [1] is also working in this area and getting better and better. The inference time is still much worse than any of the more traditional models (xg/catboost etc). [1] https://github.com/PriorLabs/tabpfn https://github.com/PriorLabs/tabpfn