Is Fitness Inheritance Useful for Real-World Applications?

Abstract

Abstract. Fitness evaluation in real-world applications often causes a lot of computational overhead. Fitness inheritance has been proposed for tackling this problem. Instead of evaluating each individual, a certain percentage of the individuals is evaluated indirectly by interpolating the fitness of their parents. However, the problems on which fitness inheritance has been tested are very simple and the question arises whether fitness inheritance is really useful for real-world applications. The objective of this paper is to test the performance of average and weighted average fitness inheritance on a well-known test suite of multiple objective optimization problems. These problems have been generated as to constitute a collection of test cases for genetic algorithms. Results show that fitness inheritance can only be applied to convex and continuous problems.

Efficient Query for Historical Data in Evolutionary Algorithm

“…With this method, a fitness value and associated reliability value are assigned to each new individual, and the individual is only evaluated if its reliability value is below a threshold. As Ducheyne et al (2003) [22]mentioned, in many real-world applications of evolutionary algorithms, the characteristics of the parents cannot be used as a good indicator of their offspring. The reason could be that although the vast majority of the offspring have some connections with their parents, but for each individual offspring, it is not necessarily similar to its parents, but instead being similar with the individuals in vicinity.…”

Ecosystem-oriented distributed evolutionary computing

“…There is a huge body of work and continuing research regarding theoretical approaches to evolutionary computing [5], including the extensive use of genetic algorithms for practical real-world problem solving [38]. Full use should make of the current state-of-the-art, and future developments in the area of evolutionary computing [39] in determining the optimal evolutionary computing approaches to have in the populations P of our ecosystem-oriented distributed evolutionary computing E.…”

Surrogate-Assisted Artificial Immune Systems for Expensive Optimization Problems

“…First proposed in (Smith et al, 1995), the fitness inheritance surrogate model has been applied in several problems (Bui et al, 2005;Ducheyne et al, 2003;Salami & Hendtlass, 2003;Sastry et al, 2004;Zheng et al, 1997) and algorithms (Pilato et al, 2008;Reyes-Sierra & Coello, 2005). In this method, all the individuals in the initial population have their fitness value calculated by the exact objective function evaluator.…”