Wide/shallow learning

[subversive-owl]

"Juice is a open Machine Learning Framework for hackers to build classical, deep or hybrid machine learning applications."

What should be available for classical and hybrid applications? I imagine linear/logistic regression, maybe SVM?

[drahnr]

Well, there are already some crates dealing with SVM and LR, so I'd rather make use of those than reimplementing. The integration of those merely to provide easy integration and prevent every user from having to write the same gluecode over and over.