Fuzzy box implementation in the Matlab

Note:This article is for the audience who have prior knowledge of fuzzy sets. This is just implementation of fuzzy sets in Matlab.

Software used: Matlab R2010a

Problem Statement: Food Tipping

Tip is decided on the basis of various attributes. Selected member function for attributes is triangular here. Range is selected so that membership functions don’t overlap with each other.

Input Variable:

S.No.	Attributes	Range
1.	Taste	[0 3]
	Delicious	[0 1]
	Normal	[1 2]
	Rancid	[2 3]
2.	Ambience	[0 3]
	Worst	[0 1]
	Good	[1 2]
	Excellent	[2 3]
3.	Service	[0 3]
	Slow	[0 1]
	On time	[1 2]
	Fast	[2 3]
4.	Price	[0 2]
	Reasonable	[0 1]
	Costly	[1 2]
5.	Hygiene	[0 2]
	Clean	[0 1]
	Stinky	[1 2]

Output Variable:

S.No.	Attributes	Range
1.	Tip	[0 1]
	100	[0 0.2]
	200	[0.2 0.4]
	300	[0.4 0.6]
	400	[0.6 0.8]
	500	[0.8 1]

Steps:
1.Start matlab and type fuzzy in command window.

      2. A new wizard named FIS Editor will open (Fuzzy Inference System).  

      3.Go on Edit->Add variable->Input. Do this for 4 times as we have five input variables. Change name of the input variable by editing the content of Name.

      4.Similarly edit the name of all five variables and output variable. Now save the work in file first. Go to File->Export-> File. Now save the food_tip.fis file. FIS editor will look like as follows:

      5. Double click on the taste input variable to edit the attributes of taste. Taste has three attributes namely: delicious, normal, rancid. New box will pop up named Membership Function Editor. Go to Edit->Remove All MFs. Then Edit-> Add Mfs. Small box asking for type and no of mf will pop up. Select trimf and 3 as we want to use the triangular member function and no of member function is 3. Click Ok.      

                                  

       6.Set range of taste variable as [0 3] (your choice). Edit mf1 name as delicious and its parameters. Parameters in triangular membership function is vertices of triangle. Select [0 0.5 1]. Similarly set all three membership function as required.

    7. Do it for all the variables. Close the membership function editor.

      8. In FIS Editor, go to Edit-> Rules. A new rule Editor will open. Now add rules one by one. Close when you are done.

      9. Now go to View->Rules. See output by varying inputs. First input is [0.5 2.5 2.5 0.5 0.5] = [taste=delicious ambience=excellent service=fast price=reasonable hygiene=clean] (as range of delicious in taste is [0 1] so 0.5 means delicious, similarly others are done) so the tip value is 0.9 so tip would be 500 (as range of 500 is [0.8 1]).

      10.Similarly, you can try for another inputs.

   Fuzzy toolbox implemented in the Matlab successfully.Thanks for reading!!