Event
Dr. Y.P. Chaubey, Concordia University
Friday, February 10, 2017 12:00to13:00
Room LB-921-04, Library Building Concordia, 1400 de Maisonneuve West, Montreal, QC, H3G 1M8, CA
Kernel density estimation: Bernstein polynomials and circular data.
In this talk I will give a brief preview of the kernel density estimator along with some new developments. Specifically, the approximation lemma due to Feller (see Lemma 1, 搂VII.1, Feller 1965) is used to motivate the Bernstein polynomial density estimator. Further, its generalization to multivariate density estimation and adaptation to circular density estimation will be presented. The talk will conclude with a result connecting the circular kernel density estimation with orthogonal polynomials on a unit circle and some illustrations.