TY - JOUR
T1 - PDA-SVM hybrid
T2 - A unified model for kernel-based supervised classification
AU - Kung, S. Y.
AU - Mak, Man Wai
N1 - Funding Information:
This manuscript was based on the keynote paper at PCM2009 by Kung [1]. This work benefited greatly from our research collaboration with Ms. Yuhui Luo from the Princeton University. The work was in part supported by The Hong Kong Research Grant Council, Grant No. PolyU5251/08E and PolyU5264/09E. Some of the research was conducted when S.Y. Kung was a Distinguished Visiting Professor at The University of Hong Kong.
Copyright:
Copyright 2014 Elsevier B.V., All rights reserved.
PY - 2011/10
Y1 - 2011/10
N2 - For most practical supervised learning applications, the training datasets are often linearly nonseparable based on the traditional Euclidean metric. To strive for more effective classification capability, a new and flexible distance metric has to be adopted. There exist a great variety of kernel-based classifiers, each with their own favorable domain of applications. They are all based on a new distance metric induced from a kernel-based inner-product. It is also known that classifier's effectiveness depends strongly on the distribution of training and testing data. The problem lies in that we just do not know in advance the right models for the observation data and measurement noise. As a result, it is impossible to pinpoint an appropriate model for the best tradeoff between the classifier's training accuracy and error resilience. The objective of this paper is to develop a versatile classifier endowed with a broad array of parameters to cope with various kinds of real-world data. More specifically, a so-called PDASVM Hybrid is proposed as a unified model for kernel-based supervised classification. This paper looks into the interesting relationship between existing classifiers (such as KDA, PDA, and SVM) and explains why they are special cases of the unified model. It further explores the effects of key parameters on various aspects of error analysis. Finally, simulations were conducted on UCI and biological data and their performance compared.
AB - For most practical supervised learning applications, the training datasets are often linearly nonseparable based on the traditional Euclidean metric. To strive for more effective classification capability, a new and flexible distance metric has to be adopted. There exist a great variety of kernel-based classifiers, each with their own favorable domain of applications. They are all based on a new distance metric induced from a kernel-based inner-product. It is also known that classifier's effectiveness depends strongly on the distribution of training and testing data. The problem lies in that we just do not know in advance the right models for the observation data and measurement noise. As a result, it is impossible to pinpoint an appropriate model for the best tradeoff between the classifier's training accuracy and error resilience. The objective of this paper is to develop a versatile classifier endowed with a broad array of parameters to cope with various kinds of real-world data. More specifically, a so-called PDASVM Hybrid is proposed as a unified model for kernel-based supervised classification. This paper looks into the interesting relationship between existing classifiers (such as KDA, PDA, and SVM) and explains why they are special cases of the unified model. It further explores the effects of key parameters on various aspects of error analysis. Finally, simulations were conducted on UCI and biological data and their performance compared.
KW - Error margin
KW - PDA-SVM Hybrid
KW - Perturbational discriminant analysis (PDA)
KW - SVM
KW - Unified model for supervised classifcation
KW - Weight-error-curve (WEC)
UR - http://www.scopus.com/inward/record.url?scp=84890443960&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84890443960&partnerID=8YFLogxK
U2 - 10.1007/s11265-011-0588-8
DO - 10.1007/s11265-011-0588-8
M3 - Article
AN - SCOPUS:84890443960
SN - 1939-8018
VL - 65
SP - 5
EP - 21
JO - Journal of Signal Processing Systems
JF - Journal of Signal Processing Systems
IS - 1
ER -