This paper introduces the concepts of error vector and error space, directly bound to semantics, one of the hottest topics in genetic programming. Based on these concepts, we introduce the notions of optimally aligned individuals and optimally coplanar individuals. We show that, given optimally aligned, or optimally coplanar, individuals, it is possible to construct a globally optimal solution analytically. Thus, we introduce a genetic programming framework for symbolic regression called Error Space Alignment GP (ESAGP) and two of its instances: ESAGP-1, whose objective is to find optimally aligned individuals, and ESAGP-2, whose objective is to find optimally coplanar individuals. We also discuss how to generalize the approach to any number of dimensions. Using two complex real-life applications, we provide experimental evidence that ESAGP-2 outperforms ESAGP-1, which in turn outperforms both standard GP and geometric semantic GP. This suggests that “adding dimensions” is beneficial and encourages us to pursue the study in many different directions, that we summarize in the final part of the manuscript

Ruberto, S., Vanneschi, L., Castelli, M., Silva, S. (2014). ESAGP – A semantic GP framework based on alignment in the error space. In 17th European Conference on Genetic Programming, EuroGP 2014 (pp.150-161) [10.1007/978-3-662-44303-3_13].

ESAGP – A semantic GP framework based on alignment in the error space

VANNESCHI, LEONARDO;CASTELLI, MAURO;
2014

Abstract

This paper introduces the concepts of error vector and error space, directly bound to semantics, one of the hottest topics in genetic programming. Based on these concepts, we introduce the notions of optimally aligned individuals and optimally coplanar individuals. We show that, given optimally aligned, or optimally coplanar, individuals, it is possible to construct a globally optimal solution analytically. Thus, we introduce a genetic programming framework for symbolic regression called Error Space Alignment GP (ESAGP) and two of its instances: ESAGP-1, whose objective is to find optimally aligned individuals, and ESAGP-2, whose objective is to find optimally coplanar individuals. We also discuss how to generalize the approach to any number of dimensions. Using two complex real-life applications, we provide experimental evidence that ESAGP-2 outperforms ESAGP-1, which in turn outperforms both standard GP and geometric semantic GP. This suggests that “adding dimensions” is beneficial and encourages us to pursue the study in many different directions, that we summarize in the final part of the manuscript
paper
genetic programming
English
European Conference on Genetic Programming, EuroGP 23-25 April
2014
Nicolau, M; Krawiec, K; Heywood, MI; Castelli, M; Garci-Sanchez, P; Merelo, JJ; Santos, MR; Sim, K
17th European Conference on Genetic Programming, EuroGP 2014
978-366244302-6
2014
8599
150
161
none
Ruberto, S., Vanneschi, L., Castelli, M., Silva, S. (2014). ESAGP – A semantic GP framework based on alignment in the error space. In 17th European Conference on Genetic Programming, EuroGP 2014 (pp.150-161) [10.1007/978-3-662-44303-3_13].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/62219
Citazioni
  • Scopus 24
  • ???jsp.display-item.citation.isi??? ND
Social impact