Consider implementing XGBoost

smartcorelib / smartcore

A comprehensive library for machine learning and numerical computing. The library provides a set of tools for linear algebra, numerical computing, optimization, and enables a generic, powerful yet still efficient approach to machine learning.

https://smartcorelib.org/

Apache License 2.0

698 stars 75 forks source link

Consider implementing XGBoost #238

Open Mec-iS opened 1 year ago

Mec-iS commented 1 year ago

I'm submitting a

[x] feature request.

Expected Behaviour:

smartcore should provide an implementation for gradient boosting (XGBoost) for most popular regressions:

Some resources here.
C++ bindings. It looks like there is still no pure-Rust implementation
forust-ml gradient boosting library

Let's try to find a low foot-print and feature-enabled solution to provide this great tool.

Luosuu commented 1 year ago

Hi @Mec-iS could you please share your update on this thread? Have you decided how to implement this? Or should we directly use the C++ bindings

Mec-iS commented 1 year ago

Hi,

At the moment this is not in our priority list as it doesn't seems that users are keen to have this feature. I would be happy to follow if somebody decides to take this up. First of all we need a list of use cases to define when/how users may want to use this (possibly using the existing datasets). We don't want to provide anything through bindings, there are already other libraries that do that pretty well; what put in there library has to go through some criteria that are loosely defined in CONTRIBUTING and DEVELOPERS.

If you have any idea please share.