Least Absolute Deviation Regression Model Discussion Contains Two Kinds Of Fuzzy Number

The thesis gives a detailed introduction to a fuzzy regression model, and then takes a classic regression analysis into the fuzzy condition analysis. What's more, it builds up a fuzzy regression model(FRM) and gives an example to the FRMThere are four chapters in this thesis. The first chapter shows us background of the research, including the ongoing research of home and abroad and tells us its arrangement.The second chapter introduces the basic concepts and the operation properties of fuzzy sets. Thereto, it also leads in the operation properties fuzzy number, and specifically pays attention to expression of trapezoid fuzzy number and analyzes that a trapezoid fuzzy number could be changed into a triangle fuzzy number at 1y =2y. What's more, it gives the definition of fuzzy distance, all of which lays a foundation for the coming chapters.The third part is a build-up fuzzy regression model. Starting from the classic regression analysis, this part takes it into the fuzzy condition. By the aid of least absolute deviation principle, it firstly gives an analysis on fuzzy regression using the least absolute deviation under triangle fuzzy number, therefore it contains the condition of trapezoid fuzzy number. Two different models are made up, one with precise input and fuzzy output, together with fuzzy parameters, and the other with all fuzzy.Finally it gives an example on analysis. Based on the data from Diamond, with the help of excel, it obtains a fuzzy regression formula, compared to Diamond's result, it shows that our approach is more workable. Another part shows a random generating trapezoid fuzzy number, with a good fitting to the giving numbers.