GOML2 aims to be:
- The most comprehensive ML library written in Go.
- Zero dependencies
While I aim to make this friendly for use in production, it's not an objective.
I intend for this library to be used directly through its APIs. It's interesting to think that we could also turn this into a binary for easy manipulation of models & hyperparameters as well as datasets, but that's currently not an objective.
I'm not sure when what the requirements are for checking these off, but I'll use this list as a rough timeline for development.
[ ] - Dataset manipulation
[ ] - Dataset visualization
[ ] - Regression trees
[ ] - Classification trees
[ ] - Gradient boosting
...
I'm currently working on this project solo, but if you've somehow managed to stumble across this repository, please don't hesitate to reach out to me at rkotcher at gmail.