Controlling overfitting in symbolic regression based on a bias/variance error decomposition

Authors:
Alexandros Agapitos;Anthony Brabazon;Michael O'Neill
Affiliations:
Financial Mathematics and Computation Research Cluster, Natural Computing Research and Applications Group, University College Dublin, Ireland;Financial Mathematics and Computation Research Cluster, Natural Computing Research and Applications Group, University College Dublin, Ireland;Financial Mathematics and Computation Research Cluster, Natural Computing Research and Applications Group, University College Dublin, Ireland
Venue:
PPSN'12 Proceedings of the 12th international conference on Parallel Problem Solving from Nature - Volume Part I
Year:
2012

Citing 10
Cited 1

The Effect of Extensive Use of the Mutation Operator on Generalization in Genetic Programming Using Sparse Data Sets

PPSN IV Proceedings of the 4th International Conference on Parallel Problem Solving from Nature
Genetic Programming, Ensemble Methods and the Bias/Variance Tradeoff - Introductory Investigations

Proceedings of the European Conference on Genetic Programming
Pattern Recognition and Machine Learning (Information Science and Statistics)

Pattern Recognition and Machine Learning (Information Science and Statistics)
Order of nonlinearity as a complexity measure for models generated by symbolic regression via Pareto genetic programming

IEEE Transactions on Evolutionary Computation
Improving symbolic regression with interval arithmetic and linear scaling

EuroGP'03 Proceedings of the 6th European conference on Genetic programming
A Field Guide to Genetic Programming

A Field Guide to Genetic Programming
Evolutionary learning of technical trading rules without data-mining bias

PPSN'10 Proceedings of the 11th international conference on Parallel problem solving from nature: Part I
A gaussian groundplan projection area model for evolving probabilistic classifiers

Proceedings of the 13th annual conference on Genetic and evolutionary computation
Maximum margin decision surfaces for increased generalisation in evolutionary decision tree learning

EuroGP'11 Proceedings of the 14th European conference on Genetic programming
A preliminary investigation of overfitting in evolutionary driven model induction: implications for financial modelling

EvoApplications'11 Proceedings of the 2011 international conference on Applications of evolutionary computation - Volume Part II

Adaptive distance metrics for nearest neighbour classification based on genetic programming

EuroGP'13 Proceedings of the 16th European conference on Genetic Programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

We consider the fundamental property of generalisation of data-driven models evolved by means of Genetic Programming (GP). The statistical treatment of decomposing the regression error into bias and variance terms provides insight into the generalisation capability of this modelling method. The error decomposition is used as a source of inspiration to design a fitness function that relaxes the sensitivity of an evolved model to a particular training dataset. Results on eight symbolic regression problems show that new method is capable on inducing better-generalising models than standard GP for most of the problems.