Multi-scale Support Vector Machine Optimization by Kernel Target-Alignment

Research areas:

• Methodology - Machine learning

Year:

2013

Type of Publication:

In Proceedings

Authors:

• Pérez-Ortiz, María
• Gutiérrez, Pedro Antonio
• Sánchez-Monedero, Javier
• Hervás-Martínez, César

Book title:

21st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN2013)

Pages:

391-396

Address:

Brudge

Organization:

Bruges, Belgium

Month:

24th-25th April

ISBN:

978-2-87419-081-0

BibTex:

@conference{MSKTA2013, author = "Mar{\'i}a P{\'e}rez-Ortiz and Pedro Antonio Guti{\'e}rrez and Javier S{\'a}nchez-Monedero and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "The problem considered is the optimization of a multi-scale kernel, where a different width is chosen for each feature. This idea has been barely studied in the literature, and through the use of evolutionary or gradient descent approaches, which explicitly train the learning machine and thereby incur high computacional cost. To cope with this limitation, the problem is explored by making use of an analytical methodology known as kernel-target alignment, where the kernel is optimized by aligning it to the so-called ideal kernel matrix. The results show that the proposal leads to better performance and simpler models at limited computational cost when applying the binary Support Vector Machine (SVM) paradigm. ", address = "Brudge", booktitle = "21st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN2013)", doi = "http://www.i6doc.com/en/livre/?GCOI=28001100131010", isbn = "978-2-87419-081-0", month = "24th-25th April", organization = "Bruges, Belgium", pages = "391-396", title = "{M}ulti-scale {S}upport {V}ector {M}achine {O}ptimization by {K}ernel {T}arget-{A}lignment", url = "https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2013-21.pdf", year = "2013", }

Abstract:

The problem considered is the optimization of a multi-scale kernel, where a different width is chosen for each feature. This idea has been barely studied in the literature, and through the use of evolutionary or gradient descent approaches, which explicitly train the learning machine and thereby incur high computacional cost. To cope with this limitation, the problem is explored by making use of an analytical methodology known as kernel-target alignment, where the kernel is optimized by aligning it to the so-called ideal kernel matrix. The results show that the proposal leads to better performance and simpler models at limited computational cost when applying the binary Support Vector Machine (SVM) paradigm.