Convex Loss Applied to Design in Regression Problems
Journal of the Royal Statistical Society: Series B (Statistical Methodology) · 1972 · Volume 34 · Issue 2
DOI: 10.1111/j.2517-6161.1972.tb00897.x
Abstract
A general linear regression function is to be observed atnpoints in order to estimate a known linear combination of the unknown parameters. Thenpoints and the estimator are to be optimum in some sense and in this paper the main criterion for optimality involves uniformly minimizing certain convex loss functions. Three main sets of results are obtained, followed by some further results for normally distributed errors.
Keywords
linear regression, design, unbiased estimates, convex loss, convex hull, tchebychev points, lagrange polynomials
Enable JavaScript for citation tools and additional paper interactions.