TY - GEN
T1 - Controlling false alarms with support vector machines
AU - Davenport, Mark A.
AU - Baraniuk, Richard G.
AU - Scott, Clayton D.
PY - 2006
Y1 - 2006
N2 - We study the problem of designing support vector classifiers with respect to a Neyman-Pearson criterion. Specifically, given a user-specified level α ε (0,1), how can we ensure a false alarm rate no greater than α while minimizing the miss rate? We examine two approaches, one based on shifting the offset of a conventionally trained SVM and the other based on the introduction of classspecific weights. Our contributions include a novel heuristic for improved error estimation and a strategy for efficiently searching the parameter space of the second method. We also provide a characterization of the feasible parameter set of the 2ν-SVM on which the second approach is based. The proposed methods are compared on four benchmark datasets.
AB - We study the problem of designing support vector classifiers with respect to a Neyman-Pearson criterion. Specifically, given a user-specified level α ε (0,1), how can we ensure a false alarm rate no greater than α while minimizing the miss rate? We examine two approaches, one based on shifting the offset of a conventionally trained SVM and the other based on the introduction of classspecific weights. Our contributions include a novel heuristic for improved error estimation and a strategy for efficiently searching the parameter space of the second method. We also provide a characterization of the feasible parameter set of the 2ν-SVM on which the second approach is based. The proposed methods are compared on four benchmark datasets.
UR - http://www.scopus.com/inward/record.url?scp=33947673977&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33947673977&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:33947673977
SN - 142440469X
SN - 9781424404698
T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
SP - V589-V592
BT - 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing - Proceedings
T2 - 2006 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2006
Y2 - 14 May 2006 through 19 May 2006
ER -