User-centric optimization with evolutionary and memetic systems

Abstract

One of the lessons learned in the last years in the metaheuristics community, and most prominently in the area of evolutionary computation (EC), is the need of exploiting problem knowledge in order to come up with effective optimization tools. This problem-knowledge can be provided in a variety of ways, but there are situations in which endowing the optimization algorithm with this knowledge is a very elusive task. This may be the case when this problem-awareness is hard to encapsulate within a specific algorithmic description, e.g., they belong more to the space of human-expert's intuition than elsewhere. An extreme case of this situation can take place when the evaluation itself of solutions is not algorithmic, but needs the introduction of a human to critically assess the quality of solutions. The above use of a combined human-user/evolutionary-algorithm approach is commonly termed interactive EC. The term user-centric EC is however more appropriate since it hints possibilities for the system to be proactive rather than merely interactive, i.e., to anticipate some of the user behavior and/or exhibit some degree of creativity. Such features constitute ambitious goals that require a good grasp of the basic underlying issues surrounding interactive optimization. An overview of these is presented in this paper, along with some hints on what the future may bring to this area. An application example is provided in the context of the search for Optimal Golomb Rulers, a very hard combinatorial problem. © 2012 Springer-Verlag.