I’ve taken a shot at developing a boosted trees package and would be interested in having feedback on its usability: https://github.com/Evovest/EvoTrees.jl
I’m conscious there’s been some initiatives in that direciton already, it was meant as a learning experiment on developing algo from scratch in the language.
Also, I wanted a design that relied on basic data structures (vanilla matrix), and easily support the hacking of new loss functions or learning paradigms.
It uses histogram methods and speed compares favourably to xgboost (at least on my uses cases).
Linear, logistic, Poisson and Quantile regressions are supported at the moment. Integration with MLJ is on its way.