ABSTRACT
We use affine arithmetic to improve both the performance and the robustness of genetic programming for symbolic regression. During evolution, we use affine arithmetic to analyze expressions generated by the genetic operators, estimating their output range given the ranges of their inputs over the training data. These estimated output ranges allow us to discard trees that contain asymptotes as well as those whose output is too far from the desired output range determined by the training instances. We also perform linear scaling of outputs before fitness evaluation. Experiments are performed on 15 problems, comparing the proposed system with a baseline genetic programming system with protected operators, and with a similar system based on interval arithmetic. Results show that integrating affine arithmetic with an implementation of standard genetic programming reduces the number of fitness evaluations during training and improves generalization performance, minimizes overfitting, and completely avoids extreme errors of unseen test data.
- J. Comba and J. Stolfi. Affine arithmetic and its applications to computer graphics. In Anais do VI Simpósio Brasileiro de Computação Gráfica e Processamento de Imagens (SIBGRAPI), 1993.Google Scholar
- L. H. de Figueiredo and J. Stolfi. Self-Validated Numerical Methods and Applications. Brazilian Mathematics Colloquium monographs. IMPA/CNPq, Rio de Janeiro, Brazil, 1997.Google Scholar
- L. H. de Figueiredo and J. Stolfi. Affine arithmetic: Concepts and applications. Numerical Algoritms, 37:147--158, 2004.Google ScholarCross Ref
- O. Gay. Libaffa - C++ affine arithmetic library for GNU/Linux. http://www.nongnu.org/libaffa/, 2006.Google Scholar
- M. Keijzer. Improving symbolic regression with interval arithmetic and linear scaling. In EuroGP: European conference on genetic programming, 2003. Google ScholarDigital Library
- M. Keijzer. Personal communication, 2010.Google Scholar
- M. Kotanchek, G. Smits, and E. Vladislavleva. Trustable symbolic regression models: using ensembles, interval arithmetic and pareto fronts to develop robust and trust-aware models. In R. Riolo, T. Soule, and B. Worzel, editors, Genetic Programming Theory and Practice V. Springer, 2008.Google ScholarCross Ref
- J. R. Koza. Genetic Programming: On the Programming of Computers by Means of Natural Selection. MIT Press, 1992. Google ScholarDigital Library
- W. B. Langdon and R. Poli. Foundations of Genetic \balancecolumns Programming. Springer-Verlag, 2002. Google ScholarDigital Library
- M. Looks. PLOP: Probabilistic Learning of Programs. http://code.google.com/p/plop, 2010.Google Scholar
- H. Majeed and C. Ryan. A less destructive, context-aware crossover operator for GP. In EuroGP: European Conference on Genetic Programming, 2006. Google ScholarDigital Library
- R. Moore. Interval Analysis. Prentice Hall, 1966.Google Scholar
- J. Stolfi. LIBAA: An affine arithmetic library in C. http://www.dcc.unicamp.br/~stolfi/, 1993.Google Scholar
- G. Valigiani, C. Fonlupt, and P. Collet. Analysis of GP improvement techniques over the real-world inverse problem of ocean colour. In EuroGP: European Conference on Genetic Programming, 2004..Google ScholarCross Ref
Recommendations
Tradeoff between Approximation Accuracy and Complexity for Range Analysis using Affine Arithmetic
Digital signal processing algorithms are usually developed in floating-point arithmetic. After that floating-point to fixed-point transformation is performed to implement them on fixed-point devices, for higher speed, smaller area and lower power. ...
Affine Arithmetic-Based B-Spline Surface Intersection with GPU Acceleration
Because the B-spline surface intersection is a fundamental operation in geometric design software, it is important to make the surface intersection operation robust and efficient. As is well known, affine arithmetic is robust for calculating the surface ...
Developing Postfix-GP Framework for Symbolic Regression Problems
ACCT '15: Proceedings of the 2015 Fifth International Conference on Advanced Computing & Communication TechnologiesThis paper describes Postfix-GP system, postfix notation based Genetic Programming (GP), for solving symbolic regression problems. It presents an object-oriented architecture of Postfix-GP framework. It assists the user in understanding of the ...
Comments