ABSTRACT
Isolating the fitness-contribution of substructures is typically a difficult task in Genetic Programming (GP). Hence, useful substructures are lost when the overall structure (model) performs poorly. Furthermore, while crossover is heavily used in GP, it typically produces offspring models with significantly lower fitness than that of the parents. In symbolic regression, this degradation also occurs because the coefficients of an evolving model lose utility after crossover. This paper proposes isolating the fitness-contribution of various substructures and reducing the negative impact of crossover by evolving a set of features instead of monolithic models. The method then leverages multiple linear regression (MLR) to optimise the coefficients of these features. Since adding new features cannot degrade the accuracy of an MLR produced model, MLR-aided GP models can bloat. To penalise such additions, we use Adjusted R2 as the fitness function. The paper compares the proposed method with standard GP and GP with linear scaling. Experimental results show that the proposed method matches the accuracy of the competing methods within only 1/10th of the number of generations. Also, the method significantly decreases the rate of post-crossover fitness degradation.
- Cesar L. Alonso, Jose Luis Montana, and Cruz Enrique Borges. 2009. Evolution Strategies for Constants Optimization in Genetic Programming. In 21st International Conference on Tools with Artificial Intelligence, ICTAI '09. IEEE, Newark, NJ, USA, 703--707. Google ScholarDigital Library
- Peter J. Angeline. 1997. Subtree Crossover: Building Block Engine or Macro-mutation?. In Genetic Programming 1997: Proceedings of the Second Annual Conference, Koza et al (Ed.). Morgan Kaufmann, Stanford University, CA, USA, 9--17. http://ncra.ucd.ie/COMP41190/SubtreeXoverBuildingBlockorMacromutation_angeline_gp97.psGoogle Scholar
- Raja Muhammad Atif Azad and Conor Ryan. 2014. A Simple Approach to Lifetime Learning in Genetic Programming based Symbolic Regression. Evolutionary Computation 22, 2 (jul 2014), 287--317. Google ScholarDigital Library
- Wolfgang Banzhaf, Peter Nordin, Robert E. Keller, and Frank D. Francone. 1998. Genetic Programming - An Introduction; On the Automatic Evolution of Computer Programs and its Applications. Morgan Kaufmann, San Francisco, CA, USA. https://www.amazon.co.uk/Genetic-Programming-Introduction-Artificial-Intelligence/dp/155860510XGoogle ScholarDigital Library
- Shu-Heng Chen and Tzu-Wen Kuo. 2003. Overfitting or Poor Learning: A Critique of Current Financial Applications of GP. In Genetic Programming, Proceedings of EuroGP'2003 (LNCS), Ryan et al (Ed.), Vol. 2610. Springer-Verlag, Essex, 34--46. Google ScholarCross Ref
- Hammad Majeed, Conor Ryan, and R. Muhammad Atif Azad. 2005. Evaluating GP schema in context. In Proceedings of the 7th annual conference on Genetic and evolutionary computation. ACM, New York, NY, USA, 1773--1774.Google Scholar
- Nicholas Freitag McPhee, Brian Ohs, and Tyler Hutchison. 2008. Semantic Building Blocks in Genetic Programming. In Proceedings of the 11th European Conference on Genetic Programming, EuroGP 2008 (Lecture Notes in Computer Science), O'Neill et al (Ed.), Vol. 4971. Springer, Naples, 134--145. Google ScholarCross Ref
Index Terms
Feature engineering for improving robustness of crossover in symbolic regression
Recommendations
Neural network crossover in genetic algorithms using genetic programming
AbstractThe use of genetic algorithms (GAs) to evolve neural network (NN) weights has risen in popularity in recent years, particularly when used together with gradient descent as a mutation operator. However, crossover operators are often omitted from ...
Exact Schema Theory and Markov Chain Models for Genetic Programming and Variable-length Genetic Algorithms with Homologous Crossover
Genetic Programming (GP) homologous crossovers are a group of operators, including GP one-point crossover and GP uniform crossover, where the offspring are created preserving the position of the genetic material taken from the parents. In this paper we ...
Comments