3
Assistant Professor, Ahvaz University Department
Abstract
Classical auditing methods such as linear and quadratic do not work well in many time series models. In this case, it is necessary to classify the time series observations in a different way. The nonparametric method, kernel auditing, is based on using the estimate of the kernel density function instead of using their actual values. The most important issue in estimating the kernel density function is choosing the appropriate value of the smoothing parameter. So far, various methods have been proposed for choosing the smoothing parameter. In this paper, various methods of estimating the kernel probability density function and the appropriate method of choosing the smoothing parameter that leads to its optimal value in time series auditing have been obtained.
