SeminarsReview of Local Polynomial Kernel Regression for Generalized Linear Models and Quasi-Likelihood Functions (Fan, J.,Heckman, N. E., and Wand, M. P. (1995))
262
reads
Generalized linear models (Nelder and Wedderburn 1972) is a useful extension of ordinary regression method. With lax assumptions, Wedderburn(1974) introduced the concept of quasi-likelihood, which presumes only the relationship of mean and variance. This paper is working on the basis of these two methods locally on each point in the range of data. To estimate each point locally, Fan, et al. induce the thought of local polynomial kernel regression. Compare to the traditional kernel regression, The new method has less the boundary effects. The asymptotic distributions of the proposed estimators are given in the text.
reads
Wei-Ying Wang
2008-11-21
13:30:00 - 15:00:00
405 , Mathematics Research Center Building (ori. New Math. Bldg.)
Generalized linear models (Nelder and Wedderburn 1972) is a useful extension of ordinary regression method. With lax assumptions, Wedderburn(1974) introduced the concept of quasi-likelihood, which presumes only the relationship of mean and variance. This paper is working on the basis of these two methods locally on each point in the range of data. To estimate each point locally, Fan, et al. induce the thought of local polynomial kernel regression. Compare to the traditional kernel regression, The new method has less the boundary effects. The asymptotic distributions of the proposed estimators are given in the text.