ARAN - Access to Research at NUI Galway

An Evolutionary Approach to Automatic Kernel Construction

ARAN - Access to Research at NUI Galway

Show simple item record

dc.contributor.author Madden, Michael G. en
dc.contributor.author Howley, Tom en
dc.date.accessioned 2009-05-15T09:57:52Z en
dc.date.available 2009-05-15T09:57:52Z en
dc.date.issued 2006 en
dc.identifier.citation "An Evolutionary Approach to Automatic Kernel Construction" , Tom Howley and Michael G. Madden. Proceedings of ICANN 2006: International Conference on Artificial Neural Networks, Athens. Lecture Notes in Computer Science (Springer), Vol. 4132, pp 417-426, Sept 2006. en
dc.identifier.uri http://hdl.handle.net/10379/190 en
dc.description.abstract Abstract. Kernel-based learning presents a unified approach to machine learning problems such as classification and regression. The selection of a kernel and associated parameters is a critical step in the application of any kernel-based method to a problem. This paper presents a data-driven evolutionary approach for constructing kernels, named KTree. An application of KTree to the Support Vector Machine (SVM) classifier is described. Experiments on a synthetic dataset are used to determine the best evolutionary strategy, e.g. what fitness function to use for kernel evaluation. The performance of an SVM based on KTree is compared with that of standard kernel SVMs on a synthetic dataset and on a number of real-world datasets. KTree is shown to outperform or match the best performance of all the standard kernels tested. en
dc.format application/pdf en
dc.language.iso en en
dc.subject Machine learning en
dc.subject Kernel based learning en
dc.subject Classification en
dc.subject Regression en
dc.subject K-Tree en
dc.subject Support vector machine (SVM) classifier en
dc.subject.lcsh Machine learning en
dc.subject.lcsh Support vector machines en
dc.title An Evolutionary Approach to Automatic Kernel Construction en
dc.type Conference Paper en

Files in this item

This item appears in the following Collection(s)

Show simple item record