The authors wish to thank Dr. Marc Zissman of Lincoln Laboratory at MIT for providing us the DARPA training and testing data. We also thank the reviewers for their valuable comments. Special thanks to Dr. Vern Paxton for his insightful comments that helped us to improve the quality and readability of the final version. This work is supported in part by the AFOSR grant F49620-01-1-0327 to the Center for Digital Security of the University of California, Davis.