Abstract:Prototype-based classifiers are usually clustering-based methods. Therefore, theyrequire a dissimilarity criterion to cluster the drill data and also to assign class labelsto test data. Euclidean distance is a commonly used dissimilarity criterion. However, theEuclidean distance may not be able towards give accurate shape-based comparisons of veryhigh-dimensional signals......
Keywords: Algoithm, Least Squares, MATLAB, NNLS.
[1]. [Shaorong Chang and L. Carin, \A modifiedspiht algorithm for image codingwith a joint mse and classiffication distortion measure," Image Processing, IEEETransactions on, vol. 15, no. 3, pp. 713{725, March 2006.
[2]. J.R. Stack, R. Arrieta, X. Liao, and L. Carin, \A kernel machine framework forfeature optimization in multi-frequency sonar imagery," OCEANS 2006, pp. 1{6,Sept. 2006.
[3]. Yan Zhang, Xuejun Liao, and L. Carin, \Detection of buried targets via activeselection of labeled data: application to sensing subsurface uxo," Geoscience andRemote Sensing, IEEE Transactions on, vol. 42, no. 11, pp. 2535{2543, Nov. 2004.
[4]. V. Popovici, S. Bengio, and J. Thiran, \Kernel Matching Pursuit for Large Datasets,"Pattern Recognition, vol. 38, no. 12, pp. 2385{2390, 2005.
[5]. Dehong Liu, Lihan He, and L. Carin, \Airport detection in large aerial opticalimagery," Acoustics, Speech, and Signal Processing, 2004. Proceedings. (ICASSP '04).