4 ms·
This is such an odd dogmatic generalization. There's plenty of business cases where e.g. SHAP is enough for explainability and having a higher accuracy is impor
by Tenoke 4y ago
This is such an odd dogmatic generalization. There's plenty of business cases where e.g. SHAP is enough for explainability and having a higher accuracy is important.
- dr_dshiv 4y agoYeah, but if you don’t know why you have high accuracy, it is hard to feel confident that your accuracy will remain high. What is SHAP?
- Tenoke 4y agoSHAP is a recent de-facto standard way to get feature importance out of any model, whether you care about how much each column in your dataset contributes to a given decision or say which pixels in an image mattered the most. https://github.com/slundberg/shap https://github.com/slundberg/shap