 |
Center for Nonlinear Studies |
Many complex real-world problems can be formulated as black-box optimization problems and the last few decades have produced a plethora of black-box optimization algorithms. Nonetheless, the actual solving of such problems by practitioners remains a significant challenge. This talk will discuss some of the challenges faced by practitioners, introduce the meta-heuristical black-box optimization tapestry metaphor, propose in this context an automated problem solving approach, and present some recent additions to the tapestry in support of the proposed approach. The additions covered will include self-configuring crossover, supportive coevolution, and the automated design of meta-heuristics employing genetic programming.