Skip to main content
Cross Validated

You are not logged in. Your edit will be placed in a queue until it is peer reviewed.

We welcome edits that make the post easier to understand and more valuable for readers. Because community members review edits, please try to make the post substantially better than how you found it, for example, by fixing grammar or adding additional resources and hyperlinks.

4
  • $\begingroup$ +1 because of mentioning Caruana, an absolute ML unit. (And obviously because it is a good answer!) $\endgroup$ Commented Jan 21, 2020 at 12:08
  • $\begingroup$ Thanks @Björn. I'm marking your answer as accepted, as I believe it covers the most relevant points here. Cheers $\endgroup$ Commented Jan 21, 2020 at 19:40
  • 2
    $\begingroup$ I think one thing that is missed in this answer (unless I misunderstood) is that perhaps the model that LIME and SHAP are trying to explain is just plain wrong. LIME and SHAP try to have features that make sense with the model, but perhaps the model is wrong. $\endgroup$ Commented Jul 13, 2021 at 11:20
  • $\begingroup$ Sure, that's also a possibility. Some of my options are particular variations of wrong models that are particularly common, but all sorts of wrong models could lead to weird explanations (and predictions). $\endgroup$ Commented Jul 13, 2021 at 12:29