Hi Guillaume,
Thanks for the sharing. Talking about ML-powered GAM, I have found two interesting approaches.
- InterpretML from Microsoft
- Mixing GAM with GBM
- Pre-built visualization methods for “black-box” model
- GAMI-net by HKU scholars
https://github.com/ZebinYang/gaminet
- Mixing GAM with NN
- The code base is not as user-friendly as Interpret-ML
I think these worth a look for the practitioners who are looking for ML-powered GAM with interpretability.