Evolution Strategy and Hierarchical Clustering

O. Aichholzer, F. Aurenhammer, B. Brandtstätter, H. Krasser, C. Magele, M. Mühlmann, and W. Renhart

Abstract:

Multi-objective optimization problems, in general, exhibit several local optima besides a global one. A desirable feature of any optimization strategy would therefore be to supply the user with as many information as possible about local optima on the way to the global solution. In this paper a hierarchical clustering algorithm implemented into a higher order Evolution Strategy is applied to achieve these goals.

Reference: O. Aichholzer, F. Aurenhammer, B. Brandtstätter, H. Krasser, C. Magele, M. Mühlmann, and W. Renhart. Evolution strategy and hierarchical clustering. In $13^{th}$ COMPUMAG Conference on the Computation of Electromagnetic Fields, Lyon-Evian, France, 2001.