I would like to request the implementation of the Simplicial Homology Global Optimisation (SHGO) algorithm by Stefan Endres.
It is a highly efficient derivative-free method for global optimization, usually requiring just a small amount of funcions evalutions. In some tests with 8-dimension Rosenbrock and Ackley functions, its Scipy implementation requires, respectively, about 750 and 360 functions evaluations, while pygmo's CMA-ES requires, respectively, about 7700 and 5090 evaluations.
Hello,
I would like to request the implementation of the Simplicial Homology Global Optimisation (SHGO) algorithm by Stefan Endres. It is a highly efficient derivative-free method for global optimization, usually requiring just a small amount of funcions evalutions. In some tests with 8-dimension Rosenbrock and Ackley functions, its Scipy implementation requires, respectively, about 750 and 360 functions evaluations, while pygmo's CMA-ES requires, respectively, about 7700 and 5090 evaluations.
Reference: https://stefan-endres.github.io/shgo/