Density deconvolution with Laplace errors and unknown variance

We consider density deconvolution with zero-mean Laplace noise in the context of an error component regression model. adapt minimax methods Meister (2006) to allow estimation unknown variance. propose a semi-uniformly consistent estimator for ordinary-smooth target and modified "variance truncation device” provide simulation study practical guidance choice smoothness parameters density. apply restricted versions our stochastic frontier model US banks measurement daily saturated fat intake.