OCTHaGOn.jl is open source!

02 Feb 2022

After the submission of my paper with Dimitris titled Global Optimization via Optimal Decision Trees, I am releasing the implementation of the method in the repository here. Right now, it is implemented for Julia 1.5 and requires an installation of Interpretable AI, available for free under an academic license.

The code is a work in progress, and documentation is pending. I encourage you to consult the test and data/baron folders in the repository to see implementations to specific global optimization instances. Please feel free to post issues if you are having problems and encounter bugs.

I am also in the process of developing code for the MIO of decision trees in the repository OptimalTrees.jl. While this is very much a work in progress, feel free to check it out for a basic training example of a globally optimal classification tree!

Berk