https://ericschmidt.xyz/tags/nanotabpfn/ Recent content in Nanotabpfn on Datamuncher Hugo -- 0.161.1 en-GB Fri, 09 Oct 2026 08:49:13 +0200 https://ericschmidt.xyz/posts/tabular-foundation-models/ Fri, 09 Oct 2026 08:49:13 +0200 https://ericschmidt.xyz/posts/tabular-foundation-models/ <h1 id="intro">Intro</h1> <p>I have been working with models for tabular data a bit professionally and as a hobby. I did notice attempts over time to use neural nets to somehow beat decision tree based models. However, they all pretty much <a href="https://proceedings.neurips.cc/paper_files/paper/2022/hash/0378c7692da36807bdec87ab043cdadc-Abstract.html">at most were on par with decision tree based models</a>, even in regimes with millions of samples.</p> <p>A recent post by <a href="https://gael-varoquaux.info/about.html">Gaël Varoquaux</a> caught my attention stating <a href="https://www.linkedin.com/posts/gael-varoquaux-a8391411_open-science-is-powering-tabular-foundation-activity-7470127515051819009-gNDU">&ldquo;Tabular foundation models have gone from a promising idea to a production-ready technology&rdquo;</a>. Notable, as he is one of the creators of <a href="http://scikit-learn.org/stable/"><code>scikit-learn</code></a> - a tool I consider an essential work horse of data science.</p>