Estimating a Convex Function in Nonparametric Regression
A new nonparametric estimate of a convex regression function is proposed and its stochastic properties are studied. The method starts with an unconstrained estimate of the derivative of the regression function, which is firstly isotonized and then integrated. We prove asymptotic normality of the new estimate and show that it is first order asymptotically equivalent to the initial unconstrained estimate if the regression function is in fact convex. If convexity is not present, the method estimates a convex function whose derivative has the same "L"-super-"p"-norm as the derivative of the (non-convex) underlying regression function. The finite sample properties of the new estimate are investigated by means of a simulation study and it is compared with a least squares approach of convex estimation. The application of the new method is demonstrated in two data examples. Copyright 2007 Board of the Foundation of the Scandinavian Journal of Statistics..
Year of publication: |
2007
|
---|---|
Authors: | BIRKE, MELANIE ; DETTE, HOLGER |
Published in: |
Scandinavian Journal of Statistics. - Danish Society for Theoretical Statistics, ISSN 0303-6898. - Vol. 34.2007, 2, p. 384-404
|
Publisher: |
Danish Society for Theoretical Statistics Finnish Statistical Society Norwegian Statistical Association Swedish Statistical Association |
Saved in:
freely available
Saved in favorites
Similar items by person
-
Estimating a convex function in nonparametric regression
Dette, Holger, (2005)
-
A note on estimating a monotone regression by combining kernel and density estimates
Dette, Holger, (2005)
-
Testing strict monotonicity in nonparametric regression
Birke, Melanie, (2006)
- More ...