Output details
11 - Computer Science and Informatics
Bangor University
Constituent Grammatical Evolution
<22>This paper was presented at the IJCAI conference, one of the two top-ranked AI conferences (acceptance rate for oral paper was 17%, 227/1325). The work describes a new evolutionary algorithm (CGE) based on Grammatical Evolution which introduces constituent genes and conditional behaviour switching. The experimental results show that CGE is able to find state-of-the-art solutions for benchmarking problems such as the Santa Fe and Los Altos Hills trails. The results also show that CGE significantly outperforms GE in terms of both efficiency (percent of solutions found) and effectiveness (number of required steps of solutions found) on all the benchmarking problems.