Building ANFIS using a toolbox

1. Building ANFIS using a toolbox Praktikum 14

2. Data • Boston Housing Data • From StatLib library at Carnegie Mellon University • Harrison, D. and Rubinfeld, D.L. 'Hedonic prices and the demand for clean air', J. Environ. Economics & Management, vol.5, 81-102, 1978. • 7 Juli 1993

3. Data (2) • 506 rows • 14 variable • 13 input  CRIM, ZN, INDUS, CHAS, NOX, RM, AGE, DIS, RAD, TAX, PTRATIO, B, and LSTAT • 1 output MEDV

4. Training and Checking • 450  training data  save in datatraining.dat • 56  checking data  save in datachecking.dat

5. GUI • Run MATLAB • Write anfisedit in command window, then push Enter in your keyboard

7. Input training data by select Training option at the Load Data frame, then push Load Data button. • Browse and choose datatraining.dat

9. Input checking data by select Checking option at the Load Data frame, then push Load Data button. • Browse and choose datachecking.dat

11. Choose Sub.Clusteringoption at the Generate FISframe, then push Generate FISbutton. • In the parameter for clustering genfisdialog box, set this value : • Range of influence : 0.5 • Squah factor : 1.25 • Accept ratio : 0.5 • Reject ratio : 0.15

12. Choose hybrid method • Error Tolerance = 0 , • Epoch = 10, then push Train Nowbutton .

13. Epoch = 10 , error = 2.0809

14. The Structure

15. Choose training dataoption at Test FISframe, then Test Now

16. Choose checking dataoption at Test FISframe, then Test Now

17. Surface