SeminarsIdentify Design Conditions to Achieve the Optimal Rate of Convergence in a Bivariate Additive Model
187
reads
This paper considers the estimation problem in a bivariate additive model, for . Here and is a random error and is assumed to be smooth which is approximated by B-splines. However, is not necessary to be smooth. This problem is motivated by the normalization needed in correcting intensity effects in microarray study. In this talk, an almost necessary and sufficient condition is given to guarantee that can be estimated with the usual one-dimensional optimal rate of convergence. We also present a simulation study to illustrate various convergence rate can be obtained when the proposed design condition is not satisfied.
reads
2008-01-15
13:30:00 - 15:00:00
405 , Mathematics Research Center Building (ori. New Math. Bldg.)
This paper considers the estimation problem in a bivariate additive model, for . Here and is a random error and is assumed to be smooth which is approximated by B-splines. However, is not necessary to be smooth. This problem is motivated by the normalization needed in correcting intensity effects in microarray study. In this talk, an almost necessary and sufficient condition is given to guarantee that can be estimated with the usual one-dimensional optimal rate of convergence. We also present a simulation study to illustrate various convergence rate can be obtained when the proposed design condition is not satisfied.