A carregar...
Publicação
A multiple expression alignment framework for genetic programming
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 464.05 KB | Adobe PDF |
Orientador(es)
Resumo(s)
Alignment in the error space is a recent idea to exploit semantic awareness in genetic programming. In a previous contribution, the concepts of optimally aligned and optimally coplanar individuals were introduced, and it was shown that given optimally aligned, or optimally coplanar, individuals, it is possible to construct a globally optimal solution analytically. As a consequence, genetic programming methods, aimed at searching for optimally aligned, or optimally coplanar, individuals were introduced. In this paper, we critically discuss those methods, analyzing their major limitations and we propose new genetic programming systems aimed at overcoming those limitations. The presented experimental results, conducted on four real-life symbolic regression problems, show that the proposed algorithms outperform not only the existing methods based on the concept of alignment in the error space, but also geometric semantic genetic programming and standard genetic programming.
Descrição
Vanneschi, L., Scott, K., & Castelli, M. (2018). A multiple expression alignment framework for genetic programming. In M. Castelli, L. Sekanina, M. Zhang, S. Cagnoni, & P. García-Sánchez (Eds.), Genetic Programming: 21st European Conference, EuroGP 2018, Proceedings, pp. 166-183. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10781 LNCS). Springer Verlag. DOI: 10.1007/978-3-319-77553-1_11
Palavras-chave
Theoretical Computer Science General Computer Science
Contexto Educativo
Citação
Editora
Springer Verlag