Brandon Greenwell comments

Results 118 comments of


                                            Brandon Greenwell

Case-specific variable importance scores

Or even a new function (e.g., `reason_codes()`)!

Conditional permutation importance

Can start with the approach taken in the **varImp** package: https://github.com/cran/varImp

Groupwise permutation importance?

Prespecified. As in all features fall into a particular group/bucket.

Groupwise permutation importance?

Example for a model with five features: x1, x2, x3, x4, and x5 ```r feature_groups

Groupwise permutation importance?

```r # This would work without having to change any other code permute_columns

Groupwise permutation importance?

I want to implement this, but is there any literature on this?!?!

Groupwise permutation importance?

Yes @ptaconet, this is exactly what I’ve been looking for, thanks for posting it!! My only struggle was whether it makes more sense to permute features in a group independently...

Proposition for new feature : Add option to compute grouped variable importance

Thanks for the suggestion @ptaconet, we’ve wondered about this for a while now: https://github.com/koalaverse/vip/issues/62. I’ll look into it!

Proposition for new feature : Add option to compute grouped variable importance

Wonder if this could also be useful in quantifying the importance of interaction effects?!

yardstick integration

The more metrics the merrier! Would you be able to put together a PR for this too 😁? If not, I could probably get around to it later this month.