Cross-task code reuse in genetic programming applied to visual learning

Open access

Abstract

We propose a method that enables effective code reuse between evolutionary runs that solve a set of related visual learning tasks. We start with introducing a visual learning approach that uses genetic programming individuals to recognize objects. The process of recognition is generative, i.e., requires the learner to restore the shape of the processed object. This method is extended with a code reuse mechanism by introducing a crossbreeding operator that allows importing the genetic material from other evolutionary runs. In the experimental part, we compare the performance of the extended approach to the basic method on a real-world task of handwritten character recognition, and conclude that code reuse leads to better results in terms of fitness and recognition accuracy. Detailed analysis of the crossbred genetic material shows also that code reuse is most profitable when the recognized objects exhibit visual similarity.

If the inline PDF is not rendering correctly, you can download the PDF file here.

  • Bhanu B. Lin Y. and Krawiec K. (2005). Evolutionary Synthesisof Pattern Recognition Systems Springer-Verlag New York NY.

  • Caruana R. (1997). Multitask learning Machine Learning28(1): 41-75.

  • Chang Y.F. Lee J.C. Mohd Rijal O. and Syed Abu Bakar S.A.R. (2010). Efficient online handwritten Chinese character recognition system using a two-dimensional functional relationship model International Journal of AppliedMathematics and Computer Science 20(4): 727-738 DOI: 10.2478/v10006-010-0055-x.

  • Ciresan D.C. Meier U. and Schmidhuber J. (2012). Multi-column deep neural networks for image classification Proceedings of the 2012 IEEE Conferenceon Computer Vision and Pattern Recognition(CVPR) Providence RI USA pp. 3642-3649.

  • Fabijańska A. (2012). A survey of subpixel edge detection methods for images of heat-emitting metal specimens International Journal of Applied Mathematicsand Computer Science 22(3): 695-710 DOI: 10.2478/v10006-012-0052-3.

  • Galvan Lopez E. Poli R. and Coello Coello C.A. (2004). Reusing code in genetic programming in M. Keijzer U.-M. O’Reilly S.M. Lucas E. Costa and T. Soule (Eds.) Genetic Programming-7th European ConferenceEuroGP 2004 Proceedings Lecture Notes in Computer Science Vol. 3003 Springer-Verlag Berlin/Heidelberg pp. 359-368.

  • Ghosn J. and Bengio Y. (2003). Bias learning knowledge sharing IEEE Transactions on Neural Networks14(4): 748-765.

  • Hall M. Frank E. Holmes G. Pfahringer B. Reutemann P. and Witten I.H. (2009). The Weka data mining software: An update SIGKDD Explorations 11(1): 10-18.

  • Haynes T. (1997). On-line adaptation of search via knowledge reuse in J.R. Koza K. Deb M. Dorigo D.B. Fogel M. Garzon H. Iba and R.L. Riolo (Eds.) GeneticProgramming 1997: Proceedings of the Second AnnualConference Morgan Kaufmann San Francisco CA pp. 156-161.

  • Holland J. (1975). Adaptation in Natural and Artificial Systems Vol. 1 University of Michigan Press Ann Arbor MI.

  • Hornby G.S. and Pollack J.B. (2002). Creating high-level components with a generative representation for body-brain evolution Artificial Life 8(3): 223-246.

  • Howard D. (2003). Modularization by multi-run frequency driven subtree encapsulation in R.L. Riolo and B. Worzel (Eds.) Genetic Programming Theory and Practice Kluwer New York NY Chapter 10 pp. 155-172.

  • Howard D. Roberts S.C. and Ryan C. (2006). Pragmatic genetic programming strategy for the problem of vehicle detection in airborne reconnaissance Pattern RecognitionLetters 27(11): 1275-1288.

  • Hsu W.H. Harmon S.J. Rodriguez E. and Zhong C. (2004). Empirical comparison of incremental reuse strategies in genetic programming for keep-away soccer in M. Keijzer (Ed.) Late Breaking Papers at the 2004 Genetic andEvolutionary Computation Conference Association for Computing Machinery Seattle WA.

  • Jaśkowski W. Krawiec K. and Wieloch B. (2007a). Genetic programming for cross-task knowledge sharing in D. Thierens (Ed.) Genetic and Evolutionary ComputationConference GECCO Association for Computing Machinery London pp. 1620-1627.

  • Jaśkowski W. Krawiec K. and Wieloch B. (2007b). Knowledge reuse in genetic programming applied to visual learning in D. Thierens (Ed.) Genetic and EvolutionaryComputation Conference GECCO Association for Computing Machinery London pp. 1790-1797.

  • Jaśkowski W. Krawiec K. and Wieloch B. (2007c). Learning and recognition of hand-drawn shapes using generative genetic programming in M. Giacobini (Ed.) EvoWorkshops2007 Lecture Notes in Computer Science Vol. 4448 Springer-Verlag Berlin/Heidelberg pp. 281-290.

  • Koza J. (1992). Genetic Programming MIT Press Cambridge MA.

  • Koza J.R. (1994). Genetic Programming II: Automatic Discoveryof Reusable Programs MIT Press Cambridge MA.

  • Koza J.R. Bennett III F.H. Andre D. and Keane M.A. (1996). Reuse parameterized reuse and hierarchical reuse of substructures in evolving electrical circuits using genetic programming in T. Higuchi (Ed.) Proceedings of InternationalConference on Evolvable Systems: From Biology toHardware (ICES-96) Lecture Notes in Computer Science Vol. 1259 Springer-Verlag Berlin.

  • Krawiec K. (2006). Learning high-level visual concepts using attributed primitives and genetic programming in F. Rothlauf (Ed.) EvoWorkshops 2006 Lecture Notes in Computer Science Vol. 3907 Springer-Verlag Berlin/Heidelberg pp. 515-519.

  • Krawiec K. (2007). Generative learning of visual concepts using multiobjective genetic programming Pattern RecognitionLetters 28(16): 2385-2400.

  • Krawiec K. and Bhanu B. (2005). Visual learning by coevolutionary feature synthesis IEEE Transactions onSystem Man and Cybernetics Part B 35(3): 409-425.

  • Kurashige K. Fukuda T. and Hoshino H. (2003). Reusing primitive and acquired motion knowledge for gait generation of a six-legged robot using genetic programming Journal of Intelligent and Robotic Systems38(1): 121-134.

  • Langdon W.B. and Poli R. (2002). Foundations of Genetic Programming Springer-Verlag New York NY.

  • Li B. Li X. Mabu S. and Hirasawa K. (2012). Towards automatic discovery and reuse of subroutines in variable size genetic network programming in X. Li (Ed.) Proceedingsof the 2012 IEEE Congress on Evolutionary ComputationBrisbane Australia pp. 485-492.

  • Louis S. and McDonnell J. (2004). Learning with case-injected genetic algorithms IEEE Transactions on EvolutionaryComputation 8(4): 316-328.

  • Luke S. (2002). ECJ evolutionary computation system http://cs.gmu.edu/eclab/projects/ecj/.

  • Mitchell T.M. (2006). The discipline of machine learning Technical Report CMU-ML-06-108 Machine Learning Department Carnegie Mellon University Pittsburgh PA.

  • Montana D.J. (1993). Strongly typed genetic programming BBN Technical Report #7866 Bolt Beranek and Newman Inc. Cambridge MA.

  • Moya M.R. Koch M.W. and Hostetler L.D. (1993). One-class classifier networks for target recognition applications World Congress on Neural Networks Portland OR USA pp. 797-801.

  • O’Sullivan J. and Thrun S. (1995). A Robot That ImprovesIts Ability to Learn Carnegie Mellon University School of Computer Science Pittsburgh PA.

  • Poli R. Langdon W.B. and McPhee I.N.F. (2008). A field guide to genetic programming http://www.gp-field-guide.org.uk/.

  • Perez C.B. and Olague G. (2013). Genetic programming as strategy for learning image descriptor operators IntelligentData Analysis 17(4): 561-583.

  • Pratt L.Y. Mostow J. and Kamm C.A. (1991). Direct transfer of learned information among neural networks Proceedingsof the 9th National Conference on Artificial Intelligence(AAAI-91) Anaheim CA USA pp. 584-589.

  • Rizki M.M. Zmuda M.A. and Tamburino L.A. (2002). Evolving pattern recognition systems IEEE Transactionson Evolutionary Computation 6(6): 594-609.

  • Roberts S.C. Howard D. and Koza J.R. (2001). Evolving modules in genetic programming by subtree encapsulation in J.F. Miller (Ed.) Genetic Programming Proceedings ofEuroGP’2001 Lecture Notes in Computer Science Vol. 2038 Springer-Verlag Berlin pp. 160-175.

  • Rosca J.P. and Ballard D.H. (1996). Discovery of subroutines in genetic programming in P.J.Angeline and K.E.Kinnear Jr. (Eds.) Advances in Genetic Programming 2MIT Press Cambridge MA Chapter 9 pp. 177-202.

  • Seront G. (1995). External concepts reuse in genetic programming in E.V. Siegel and J.R. Koza (Eds.) WorkingNotes for the AAAI Symposium on Genetic Programming AAAI/MIT Cambridge MA pp. 94-98.

  • TackettW.A. (1993). Genetic generation of “dendritic” trees for image classification Proceedings of theWorld Congress onNeural Networks Portland OR USA pp. IV 646-649.

  • Teller A. and Veloso M. (1997). PADO: A new learning architecture for object recognition in K. Ikeuchi and M. Veloso (Eds.) Symbolic Visual Learning Oxford Press New York NY pp. 77-112.

  • Trujillo L. and Olague G. (2006). Synthesis of interest point detectors through genetic programming in M. Cattolico (Ed.) Genetic and Evolutionary Computation ConferenceGECCO Association for Computing Machinery Seattle WA pp. 887-894.

  • Vilalta R. and Drissi Y. (2002). A perspective view and survey of meta-learning Artificial Intelligence Review18(2): 77-95.

  • Whitley D. Rana S. and Heckendorn R. (1999). The island model genetic algorithm: On separability population size and convergence Journal of Computing and InformationTechnology 7(1): 33-47.

Search
Journal information
Impact Factor

IMPACT FACTOR 2018: 1.504
5-year IMPACT FACTOR: 1.553

CiteScore 2018: 2.09

SCImago Journal Rank (SJR) 2018: 0.493
Source Normalized Impact per Paper (SNIP) 2018: 1.361

Mathematical Citation Quotient (MCQ) 2018: 0.08

Cited By
Metrics
All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 169 79 2
PDF Downloads 78 52 0