Achieving interpretable machine learning by functional decomposition of black-box models into explainable predictor effects. https://t.co/otEMddQFlE

Achieving interpretable machine learning by functional decomposition of black-box models into explainable predictor effects https://t.co/jdTQpJFJPy

GM ＞|＜@0xsudogm

2 years ago

Historically, white-box ML models struggled due to technical debt and constraints. People demanded it to work, be explainable, and remain simple, often led to cascading errors. Black-box models have taken the lead partly because of a simpler philosophy: we just want it to work.