Abstract HTML Views: 164 PDF Downloads: 71 Total Views/Downloads: 235
Abstract HTML Views: 93 PDF Downloads: 44 Total Views/Downloads: 137
A computationally fast and optimally smooth method for generating a probability density of y given x that models given data points is described and illustrated. This method interpolates in that the mean function intersects the points and the variance function is zero at the points. It is fast and optimal in that it is produced by the smallest number of maximally-smooth Gaussian radial interpolators for which the extrapolated density has the mean and variance of the classic linear model.