Comparative research of estimation accuracy of parameters of fuzzy regression models with various types of m-estimates

In work is conducted a comparative research of estimation accuracy of parameters of fuzzy regression models with various types of m-estimates, are considered several types of loss known functions. The purpose of this research is to obtain practical recommendations for using the considered of loss functions for constructing robust fuzzy regression. Were considered two options for clogging the data sample: with a symmetric clogging or asymmetric clogging. Samples observations range from small 25 to 300. Was analyzed the rate norm bias estimates decrease and the mean square error with increasing number of observations with various loss functions. Make are certain conclusions.