Genetic programming: on the programming of computers by means of natural selection
Genetic programming: on the programming of computers by means of natural selection
Effects of locality in individual and population evolution
Advances in genetic programming
Advances in genetic programming
The evolution of size and shape
Advances in genetic programming
Advanced Population Diversity Measures in Genetic Programming
PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
Code growth in genetic programming
Code growth in genetic programming
Towards identifying populations that increase the likelihood of success in genetic programming
GECCO '05 Proceedings of the 7th annual conference on Genetic and evolutionary computation
Dormant program nodes and the efficiency of genetic programming
GECCO '05 Proceedings of the 7th annual conference on Genetic and evolutionary computation
On the behavioral diversity of random programs
Proceedings of the 9th annual conference on Genetic and evolutionary computation
Characterizing diversity in genetic programming
EuroGP'06 Proceedings of the 9th European conference on Genetic Programming
Diversity in genetic programming: an analysis of measures and correlation with fitness
IEEE Transactions on Evolutionary Computation
| Hi-index | 0.02 |
Promoting and maintaining diversity in a population is considered an important element of evolutionary computing systems, and genetic programpming (GP) is no exception. Diversity metrics in GP are usually based on structural program characteristics, but even when based on behaviour they almost always relate to fitness. We deviate from this in two ways: firstly, by considering an alternative view of diversity based on the actual activity performed during execution, irrespective of fitness; and secondly, by examining the effects of applying associated diversity-enhancing algorithms to the initial population only. Used together with an extension to this approach that provides for additional filtering of candidate population members, the techniques offer significant performance improvements when applied to the Santa Fe artificial ant problem and a maze navigation problem.