@inproceedings{Gagne:PPSN:2006,
	title = {Genetic Programming for Kernel-Based Learning with
                 Co-evolving Subsets Selection},
	address = {Reykjavik, Iceland},
	author = {Christian Gagne and Marc Schoenauer and Michele Sebag and Marco Tomassini},
	booktitle = {Parallel Problem Solving from Nature - PPSN IX},
	editor = {Thomas Philip Runarsson and Hans-Georg Beyer and Edmund Burke and Juan J. Merelo-Guervos and L. Darrell Whitley and Xin Yao},
	month = {9-13 September},
	pages = {1008--1017},
	publisher = {Springer-Verlag},
	series = {LNCS},
	url = {http://ppsn2006.raunvis.hi.is/proceedings/287.pdf},
	volume = {4193},
	year = {2006},
	biburl = {http://www.bibsonomy.org/bibtex/2eef5a5cc1ab76a498f137868094afb73/brazovayeye},
	abstract = {Support Vector Machines (SVMs) are well-established
                 Machine Learning (ML) algorithms. They rely on the fact
                 that i) linear learning can be formalised as a
                 well-posed optimisation problem; ii) nonlinear learning
                 can be brought into linear learning thanks to the
                 kernel trick and the mapping of the initial search
                 space onto a high dimensional feature space. The kernel
                 is designed by the ML expert and it governs the
                 efficiency of the SVM approach. In this paper, a new
                 approach for the automatic design of kernels by Genetic
                 Programming, called the Evolutionary Kernel Machine
                 (EKM), is presented. EKM combines a well-founded
                 fitness function inspired from the margin criterion,
                 and a co-evolution framework ensuring the computational
                 scalability of the approach. Empirical validation on
                 standard ML benchmark demonstrates that EKM is
                 competitive using state-of-the-art SVMs with tuned
                 hyper-parameters.},
	publisher_address = {Berlin}, size = {10 pages}, isbn = {3-540-38990-3}, notes = {PPSN-IX

                 evolved Kernels are forced to be symmetric functions.
                 Mercer's condition not enforced, but evolved. 3
                 co-evolving populations. runtime < 1 hour. Size based
                 parsimony pressure. Comparison with k-nn nearest
                 neighbours and SVM, GK-SVM (both with somewhat
                 optimised parameters). 6 undemanding UCI benchmarks.}, doi = {doi:10.1007/11844297_102},
	keywords = {DSS, algorithms, beagle coevolution, genetic hyperheuristic, open programming, }
}