Binary interaction methods for high dimensional global optimization and machine learning
Abstract
In this work we introduce a new class of gradient-free global optimization methods based on a binary interaction dynamics governed by a Boltzmann type equation. In each interaction the particles act taking into account both the best microscopic binary position and the best macroscopic collective position. In the mean-field limit we show that the resulting Fokker-Planck partial differential equations generalize the current class of consensus based optimization (CBO) methods. For the latter methods, convergence to the global minimizer can be shown for a large class of functions. Algorithmic implementations inspired by the well-known direct simulation Monte Carlo methods in kinetic theory are derived and discussed. Several examples on prototype test functions for global optimization are reported including applications to machine learning.
- Publication:
-
arXiv e-prints
- Pub Date:
- May 2021
- DOI:
- arXiv:
- arXiv:2105.02695
- Bibcode:
- 2021arXiv210502695B
- Keywords:
-
- Mathematics - Optimization and Control
- E-Print:
- doi:10.1007/s00245-022-09836-5