Showing papers in "arXiv: Methodology in 2003"
Posted Content•
TL;DR: In this article, the authors proposed new estimates for the frontier of a set of points, which are defined as kernel estimates covering all the points and whose associated support is of smallest surface, written as linear combinatio- ns of kernel functions applied to the points of the sample.
Abstract: We propose new estimates for the frontier of a set of points. They are defined as kernel estimates covering all the points and whose associated support is of smallest surface. The estimates are written as linear combinatio- ns of kernel functions applied to the points of the sample. The coefficients of the linear combination are then computed by solving a linear programming problem. In the general case, the solution of the optimizat- ion problem is sparse, that is, only a few coefficients are non zero. The corresponding points play the role of support vectors in the statistical learning theory. The L_1 error between the estimated and the true frontiers is shown to be almost surely converging to zero, and the rate of convergence is provided. The behaviour of the estimates on finite sample situations is illustrated on some simulations.
6 citations
Posted Content•
TL;DR: In this article, the SiZer method is extended to nonparametric hazard estimation and also to censored density and hazard estimation, allowing quick, visual statistical inference about the important issue of statistically significant increases and decreases in the smooth curve estimate.
Abstract: The SiZer method is extended to nonparametric hazard estimation and also to censored density and hazard estimation. The new method allows quick, visual statistical inference about the important issue of statistically significant increases and decreases in the smooth curve estimate. Instead of being straightforward, this extension has required the opening of a new avenue of research on the interface between statistical inference and scale space.
4 citations