
Deep learning vs tabular models (Ep. 217)
Data Science at Home
00:00
Creating a Benchmark for Tabular Learning
There's no standard benchmark when it comes to tabular data. So what these group of researchers has been doing is essentially creating a benchmark for tabular learning. They have to create the conditions to run experiments and then measure certain metrics that can be specific to that experiment. Of course, the quality of the study also depends on how general or how applicable these metrics are.
Transcript
Play full episode