OCTHaGOn

OCTHaGOn (Optimal Classification Trees with Hyperplanes for Global Optimization) is a Julia package that allows for the solution of global optimization problems using mixed-integer (MI) linear and convex approximations. It is an implementation of the methods detailed in Chapter 2 of this thesis and submitted to Operations Research. OCTHaGOn is licensed under the MIT License.

OCTHaGOn leans on the JuMP.jl modeling language in its backend, and it develops MI approximations using Interpretable AI, with a free academic license. The problems can then be solved by JuMP-compatible solvers, depending on the type of approximation. OCT's default solver in tests is CPLEX, which is free with an academic license as well.

This documentation is a work in progress. If you have any burning questions or applications, or are having problems with OCTHaGOn, please create an issue!