Comments on Leo Breiman's paper 'Statistical Modeling: The Two Cultures' (Statistical Science, 2001, 16(3), 199-231)
This is an incremental commentary that critiques the statistical community's response to Breiman's ideas and calls for a shift towards inferential evaluation in machine learning.
The paper revisits Leo Breiman's critique of statistical modeling, highlighting the ongoing tension between traditional statistical inference and modern model-free machine learning, and argues that current success in prediction must now address deeper inferential questions about algorithm stability and causality.
Breiman challenged statisticians to think more broadly, to step into the unknown, model-free learning world, with him paving the way forward. Statistics community responded with slight optimism, some skepticism, and plenty of disbelief. Today, we are at the same crossroad anew. Faced with the enormous practical success of model-free, deep, and machine learning, we are naturally inclined to think that everything is resolved. A new frontier has emerged; the one where the role, impact, or stability of the {\it learning} algorithms is no longer measured by prediction quality, but an inferential one -- asking the questions of {\it why} and {\it if} can no longer be safely ignored.