An approach to fuzzy regression analysis with fuzzy data

Regression analysis that has an important role in statistical analysis methods is procedure that identify cause and effect relation between one dependent quantitative variable and one or more independent quantitative variable. Mathematical function that indicates the form of relation is also regression model. Fuzzy theory is the method which is employed for the event that data set is vague and not obvious or if there is lack of information about problems parameters. Fuzzy regression is implemented as an alternative method depending on uncertainty of event, whole or part of data set is fuzzy or if the relationship between system parameters aren't defined certainly and classical regression models aren't suggested. In this study, fuzzy set and regression analysis are investigated in theoretical background and it is proposed an alternative approach to parameter estimation of fuzzy regression method. In application part, data set which can be found in literature is used. With the help of Python 3.4 software, fuzzy parameter estimation method is supported with numerical sample. Significance of constructed model is tested with obtained analysis of variance table using WABL deffuzification method. Also fuzzy regression model is constructed in order to estimate the amount of ink, varnish and solvent inside the boiler where is found in stamping machine in a package company that carry on a business.