Estimation Of The Stochastic Frontier Model Parameters With The Nonlinear Smooth Transfer Function

Abstract

The purpose of the present study was to estimate the parameters of the new stochastic Frontier model by nonlinear smooth transfer function structure with compound error. In the stochastic Frontier model, the compound error consists of two statistical components: model error and technical inefficiency, so that the technical inefficiency of the function is assumed to be autocorrelated. For the parametric part, the nonlinear smooth transfer function was used using the Taylor series, and for the non-parametric part, the moderating factor was used to estimate the model parameters. The research model was evaluated using data on the number of patients admitted to hospitals for Covid 19 disease. The result showed that the parameter estimates are consistent and have less error than conventional models.