Thanks for the newsletter, although quite impressive, this seems like a lot of overhead for imputing missing values compared to just imputing using the mean. I guess when it is a high stakes case and the missing values are few, this makes sense. Would love to get your comments on the kind of data that where it would necessitate this or where this kind of imputing works well. Enjoyed the read
Thanks for the newsletter, although quite impressive, this seems like a lot of overhead for imputing missing values compared to just imputing using the mean. I guess when it is a high stakes case and the missing values are few, this makes sense. Would love to get your comments on the kind of data that where it would necessitate this or where this kind of imputing works well. Enjoyed the read
Does this imputation apply to categorical features?