Anthony Blaom, PhD

Results 815 comments of Anthony Blaom, PhD

A bunch of links from the tabs to intro to stats learning are broken. I think there is a spelling mistake (my bad): ("learning" not ".earning")

Perhaps I'm missing something, but I don't believe it makes sense to apply MultinomialNB to the iris data. As @dfdx implies, the input data for a MultinomialNB is categorical not...

A related issue: It would be useful if the "regularisation" parameter alpha in Multinomial case is allowed to be continuous ([Lidstone smoothing](https://en.wikipedia.org/wiki/Naive_Bayes_classifier)). Currently it must be discrete, I think. I...

Well, that would be awesome. > If I were to re-implement current algorithms in MLJModels, how would you estimate effort for someone new to MLJ infrastructure? Well, pretty easy, I...

Multitarget regression is not possible - for regression currently the target must have `Float64` type.

Copying post of @fipelle from [another thread](https://github.com/JuliaAI/DecisionTree.jl/issues/217#issuecomment-1484446032): --- Also, it seems that loss is only available for classification trees - not regression trees. Is it possible to repurpose the existing...

This is an question here and I have wondered before why classification and regression have separate implementations. I diff'ed the two `tree.jl` files and found quite a few differences. For...

I'd say we'd want this generically. `pdf` should propagate missing values.

@filipebraida I think this can be closed, no?