ÿþ<html><head> <title>Anomaly Detection Through a Bayesian Support Vector Machine</title></head> <body bgcolor="#ffffff"> <center> <i>Journal of Transaction on Reliability, Vol. 1, No. 1, pp. 1-11, June 2010 </i> <br><h2>Anomaly Detection Through a Bayesian Support Vector Machine</h2> <p><strong>Vasilis A. Sotiris </strong><br> Center for Advanced Life Cycle Engineering (CALCE) <br> University of Maryland<br> College Park, MD 20742 USA</p> <p><strong>Peter Tse </strong><br> City University in Hong Kong <br> Kowloon, Hong Kong<br> </p> <p><strong>Michael Pecht, <em>Member, IEEE</em></strong><br> City University in Hong Kong <br> Kowloon, Hong Kong<br> and <br> Center for Advanced Life Cycle Engineering (CALCE) <br> University of Maryland <br> College Park, MD 20742 USA</p> </center> <p><b>Abstract:</b></p> <p>This paper investigates the use of a one-class support vector machine algorithm to detect the onset of system anomalies, and trend output classification probabilities, as a way to monitor the health of a system. In the absence of unhealthy (negative class) information, a marginal kernel density estimate of the healthy (positive class) distribution is used to construct an estimate of the negative class. The output of the one-class support vector classifier is calibrated to posterior probabilities by fitting a logistic distribution to the support vector predictor model in an effort to manage false alarms. <p><strong>Index Terms </strong>&mdash; Support Vector machine, Bayesian linear models, one-class classifier, anomaly detection, kernel density estimation, Bayesian posterior class probabilities. <p><a href="../../fulltext/2010/Anomaly_Detection_Bayesian_Support_Vector_Machine.pdf">Complete article</a> is available to CALCE Consortium Members.</p> <p><font size="-2"><font color="red">&copy; IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.</font></font></p> <hr><br> <center> [<a href="http://www.calce.umd.edu">Home Page</a>] [<a href="../../../articles">Articles Page</a>] </center> <center> <font size="-1">Copyright &copy; 2010 by CALCE and the University of Maryland, All Rights Reserved </font> </center> </body></html>