On the Generalization Ability of Geometric Semantic Genetic Programming

详细信息    查看全文

• 作者：Ivo Gon莽alves (21) (22)
  Sara Silva (21) (22) (23)
  Carlos M. Fonseca (21)
  
  21. CISUC ; Department of Informatics Engineering ; University of Coimbra ; 3030-290 ; Coimbra ; Portugal
  22. BioISI - Biosystems and Integrative Sciences Institute ; Faculty of Sciences ; University of Lisbon ; Campo Grande ; 1749-016 ; Lisbon ; Portugal
  23. NOVA IMS ; Universidade Nova de Lisboa ; 1070-312 ; Lisbon ; Portugal
  
• 关键词：Geometric semantic genetic programming ; Generalization ; Overfitting ; Pharmacokinetics ; Drug discovery
• 刊名：Lecture Notes in Computer Science
• 出版年：2015
• 出版时间：2015
• 年：2015
• 卷：9025
• 期：1
• 页码：41-52
• 全文大小：328 KB
• 参考文献：1. Archetti, F, Lanzeni, S, Messina, E, Vanneschi, L (2007) Genetic programming for computational pharmacokinetics in drug discovery and development. Genet. Program. Evolvable Mach. 8: pp. 413-432 9040-z" target="_blank" title="It opens in new window">CrossRef
  2. Castelli, M, Silva, S, Vanneschi, L (2014) A C++ framework for geometric semantic genetic programming. Genet. Program. Evolvable Mach. 15: pp. 73-81
  3. Dietterich, TG Ensemble methods in machine learning. In: Kittler, J, Roli, F eds. (2000) Multiple Classifier Systems. Springer, Heidelberg, pp. 1-15 9_1" target="_blank" title="It opens in new window">CrossRef
  4. Gon莽alves, I, Silva, S Balancing learning and overfitting in genetic programming with interleaved sampling of training data. In: Krawiec, K, Moraglio, A, Hu, T, Etaner-Uyar, A艦, Hu, B eds. (2013) Genetic Programming. Springer, Heidelberg, pp. 73-84 978-3-642-37207-0_7" target="_blank" title="It opens in new window">CrossRef
  5. Hansen, LK, Salamon, P (1990) Neural network ensembles. IEEE Trans. Pattern Anal. Mach. Intell. 12: pp. 993-1001 9/34.58871" target="_blank" title="It opens in new window">CrossRef
  6. Koza, JR (1992) Genetic Programming: On the Programming of Computers by Means of Natural Selection (Complex Adaptive Systems). The MIT Press, Cambridge
  7. Moraglio, A.: Towards a geometric unification of evolutionary algorithms. Ph.D. thesis, Department of Computer Science, University of Essex, UK, November 2007
  8. Moraglio, A, Krawiec, K, Johnson, CG Geometric semantic genetic programming. In: Coello, CAC, Cutello, V, Deb, K, Forrest, S, Nicosia, G, Pavone, M eds. (2012) Parallel Problem Solving from Nature - PPSN XII. Springer, Heidelberg, pp. 21-31 978-3-642-32937-1_3" target="_blank" title="It opens in new window">CrossRef
  9. Poli, R., Langdon, W.B., McPhee, N.F., Koza, J.R.: A field guide to genetic programming (2008). Lulu.com
  10. Vanneschi, L, Castelli, M, Manzoni, L, Silva, S A new implementation of geometric semantic GP and its application to problems in pharmacokinetics. In: Krawiec, K, Moraglio, A, Hu, T, Etaner-Uyar, A艦, Hu, B eds. (2013) Genetic Programming. Springer, Heidelberg, pp. 205-216 978-3-642-37207-0_18" target="_blank" title="It opens in new window">CrossRef
  
• 作者单位：Genetic Programming
• 丛书名：978-3-319-16500-4
• 刊物类别：Computer Science
• 刊物主题：Artificial Intelligence and Robotics
  Computer Communication Networks
  Software Engineering
  Data Encryption
  Database Management
  Computation by Abstract Devices
  Algorithm Analysis and Problem Complexity
  
• 出版者：Springer Berlin / Heidelberg
• ISSN：1611-3349

文摘

Geometric Semantic Genetic Programming (GSGP) is a recently proposed form of Genetic Programming (GP) that searches directly the space of the underlying semantics of the programs. The fitness landscape seen by the GSGP variation operators is unimodal with a linear slope by construction and, consequently, easy to search. Despite this advantage, the offspring produced by these operators grow very quickly. A new implementation of the same operators was proposed that computes the semantics of the offspring without having to explicitly build their syntax. This allowed GSGP to be used for the first time in real-life multidimensional datasets. GSGP presented a surprisingly good generalization ability, which was justified by some properties of the geometric semantic operators. In this paper, we show that the good generalization ability of GSGP was the result of a small implementation deviation from the original formulation of the mutation operator, and that without it the generalization results would be significantly worse. We explain the reason for this difference, and then we propose two variants of the geometric semantic mutation that deterministically and optimally adapt the mutation step. They reveal to be more efficient in learning the training data, and they also achieve a competitive generalization in only a single operator application. This provides a competitive alternative when performing semantic search, particularly since they produce small individuals and compute fast.