Don’t evaluate, inherit
12 July 2001Sastry, K., Goldberg, D. E., Pelikan, M. (2001). Proceedings of the Genetic and Evolutionary Computation Conference, 551—558. [Full paper - PDF] [Full paper - PS] [Presentation slides].
Abstract:
This paper studies fitness inheritance as an efficiency enhancement technique for genetic and evolutionary algorithms. Convergence and population-sizing models are derived and compared with experimental results. These models are optimized for greatest speed-up and the optimal inheritance proportion to obtain such a speed-up is derived. Results on OneMax problems show that when the inheritance effects are considered in the population-sizing model, the number of function evaluations are reduced by 20% with the use of fitness inheritance. Results indicate that for a fixed population size, the number of function evaluations can be reduced by 70% using a simple fitness inheritance technique.
Related Posts:
- Do not match, Inherit: Fitness surrogates for genetics-based machine learning techniques
- Inherit the wind
- Substructural neighborhoods for local search in the Bayesian optimization algorithm
- Linkage learning in real-coded GAs with simplex crossover
- Do not match, inherit: Fitness surrogates for genetics-based machine learning techniques
Comments are closed.
