When the observed data are imprecise, uncertain regression model is more suitable for linear analysis. Least squares estimation can fully consider given and minimize sum of residual error effectively solve equation imprecisely data. On basis uncertainty theory, this paper presents an deformation method solving unknown parameters in equations. We first establish one-dimensional then extend it to...

We consider a spatial functional linear regression, where scalar response is related to square-integrable process. use smoothing spline estimator for the slope parameter and establish finite sample bound variance of this estimator. Then we give optimal prediction error under mixing dependence. Finally, illustrate our results by simulations an application ozone pollution forecasting at nonvisite...