v1.1.1

Updates

• Fixed a bug with the random state of the MaxMarginIntervalTree class. It is now validated at training time instead of in the constructor for better sklearn compatibility.

• Other minor improvements

v1.0.0

Python package:

• Learning Maximum Margin Interval Trees
• Minimum cost-complexity pruning
• Cross-validation (grid search)

C++ core module:

• A dynamic programming solver to minimize a sum of hinge losses
• Python and R bindings

R package:

• R language interface for the solver