On some approaches to efficient derivation of sparse representations for optimal prediction with application to classification and regression via kernel methods Ernest Fokoue, Center for Quality and Applied Statistics, RIT Tuesday, April 26 ABSTRACT We propose a novel approach to achieving sparse representation for kernel machines through a straightforward algorithm that consists in a refinement of the maximum a posteriori (MAP) estimator of the weights of the kernel expansion. Our proposed method combines structured prior matrices and functions of the information matrix to zero in on a very sparse representation. We show computationally that our naturally efficient sparsity tuner (NEST) achieves a very sparse and predictively accurate estimator of the underlying function, for a variety of choices of the covariance matrix of our Gaussian prior over the weights of the kernel expansion. Our computational comparisons on both artificial and real examples show that our method compete very well - usually favorably - with the Support Vector Machine, the Relevance Vector Machine and Gaussian Process regressors. BIOGRAPHY Ernest P. Fokoue is an Assistant Professor of Statistics with the Center for Quality and Applied Statistics at Rochester Institute of Technology. Prior to joining RIT, Dr. Fokoue was a faculty member in the Mathematics Department at Kettering University in Flint, Michigan. Before that he was an Assistant Professor at Ohio State University (still a true Buckeye fan) and a Postdoctoral Research Fellow/Visiting Assistant Professor at the Statistical and Applied Mathematical Sciences Institute (SAMSI)/Duke University. Dr. Fokoue obtained the Masters of Science degree in Neural Computation from Aston University in Birmingham, after which he moved up north to the University of Glasgow where he earned his Ph.D. in Statistics under the supervision of Professor Mike Titterington. Dr. Fokoue's main research interest is Statistical Machine Learning and Data Mining. He recently co-authored the book titled"Principles and Theory for Data Mining and Machine Learning" published by Springer-Verlag, New York. Dr. Fokoue was the winner of the Best Young Researchers' Award from the International Association of Statistical Computing in August 2000 in Utretcht - Netherlands.