Member-only story
Deep Learning Is Not ‘Worse’ than Trees on Tabular Data
Moving beyond classical conceptions of what tabular data can be
Deep learning has become the public and private face of Artificial Intelligence. When one talks casually about Artificial Intelligence with friends at a party, strangers on the street, and colleagues at work, it is almost always on the exciting models which generate language, create art, synthesize music, and so on. Massive and intricately designed deep learning models power most of these exciting machine capabilities.
Many practitioners, however, are rightfully pushing back against the technological sensationalism of deep learning. While deep learning is “what’s cool,” it certainly is not the end-all be-all of modeling.
While deep learning has undoubtedly dominated specialized, high-dimensional data forms such as images, text, and audio, the general consensus is that it performs comparatively worse in tabular data.
It is therefore tabular data where those with some distaste, or even resentment, towards deep learning stake out their argument. (It was and still is fashionable to publish accepted deep learning papers which make seemingly trivial or even scientifically dubious modifications. This is one of the gripes against deep learning research culture articulated by…
