Learnable Embeddings of Program Spaces
We consider a class of adaptive, globally-operating, semantic-based embeddings of programs into discrete multidimensional spaces termed prespaces. In the proposed formulation, the original space of programs and its prespace are bound with a learnable mapping, where the process of learning is aimed at improving the overall locality of the new representation with respect to program semantics. To learn the mapping, which is formally a permutation of program locations in the prespace, we propose two algorithms: simple greedy heuristics and an evolutionary algorithm. To guide the learning process, we use a new definition of semantic locality. In an experimental illustration concerning four symbolic regression domains, we demonstrate that an evolutionary algorithm is able to improve the embedding designed by means of greedy search, and that the learned prespaces usually offer better search performance than the original program space.
Unable to display preview. Download preview PDF.
- 3.Dijkstra, E.W.: On the cruelty of really teaching computing science. circulated privately (December 1988)Google Scholar
- 5.Krawiec, K., Lichocki, P.: Approximating geometric crossover in semantic space. In: Raidl, G., et al. (eds.) GECCO 2009: Proceedings of the 11th Annual Conference on Genetic and Evolutionary Computation (2009)Google Scholar
- 6.Looks, M.: Competent Program Evolution. Doctor of science, Washington University, St. Louis, USA (December 11, 2006)Google Scholar
- 7.Miller, J.F.: An empirical study of the efficiency of learning boolean functions using a cartesian genetic programming approach. In: Banzhaf, W., Daida, J., Eiben, A.E., Garzon, M.H., Honavar, V., Jakiela, M., Smith, R.E. (eds.) Proceedings of the Genetic and Evolutionary Computation Conference, Orlando, Florida, USA, July 13-17, vol. 2, pp. 1135–1142. Morgan Kaufmann, San Francisco (1999)Google Scholar
- 9.Nguyen, Q.U., Nguyen, X.H., O’Neill, M.: Semantic aware crossover for genetic programming: The case for real-valued function regression. In: Vanneschi, L., Gustafson, S., Moraglio, A., De Falco, I., Ebner, M. (eds.) EuroGP 2009. LNCS, vol. 5481, pp. 292–302. Springer, Heidelberg (2009)CrossRefGoogle Scholar
- 11.Rothlauf, F.: On the locality of representations. Technical Report, University of Mannheim, Department of Information Systems 1 (2003)Google Scholar
- 13.Vanneschi, L., Tomassini, M.: Pros and cons of fitness distance correlation in genetic programming. In: Barry, A.M. (ed.) GECCO 2003: Proceedings of the Bird of a Feather Workshops, Genetic and Evolutionary Computation Conference, Chigaco, July 11, pp. 284–287. AAAI, Menlo Park (2003)Google Scholar