Fuzzy regression and an econometric application

The initial purpose of the study is to define fuzzy logic which was suggested by Zadeh in 1965 and which covers a substantial part of a daily life, usually in an unconscious sense. Then, the latter purpose of the study is to explain fuzzy regression models which are developed due to fuzzy logic. While classical logic doesn't involve linguistic statements, many values are expressed as linguistically in fuzzy logic. Thus, classical logic set includes certain values, but fuzzy logic set includes possible values and also certain values. As we know, in statistics, classical regression means explaining of one dependent variable by one or more than one independent variables. But because of a reliable analysis requires certain informations from nature, estimations of classical regression may not be consistent with the real life, occasionally. In researches, it is suggested that in some cases, fuzzy regression offers significant results than classical regression and it leads to more consistent results with the real life. In this context, the study attempts to present that the results of fuzzy regression are better than the classical regression. Thus, in the first part of the econometric analysis, we set a consumption function by using the TCMB's data of quantity of consumption and GDP in 2000-2012 periods and made estimation with fuzzy regression. Then, in the second part of the econometric analysis, we set an import function by using the data of quantity of import, GDP and exchange rates and changes of the quantity of import are analysed comparatively by using both classical and fuzzy regression methods.