A First Course in FUZZY and NEURAL CONTROL

274 CHAPTER 8. APPLICATIONS

Initial MF's on Area

0

0.2

0.4

0.6

0.8

1

1.2

0 1000 2000 3000 4000 5000

Area

Final MF's on Area

0

0.2

0.4

0.6

0.8

1

1.2

0 1000 2000 3000 4000 5000

Area

Figure 8.15. Initial andfinal membership functions onarea: 232-partition

Initial MF's on Solidity

0

0.2

0.4

0.6

0.8

1

1.2

0 0.2 0.4 0.6 0.8 1

Solidity

Final MF's on Solidity

0

0.2

0.4

0.6

0.8

1

1.2

0 0.2 0.4 0.6 0.8 1

Solidity

Figure 8.16. Initial andfinal membership functions onsolidity: 232-partition

Initial MF's on Edif

0

0.2

0.4

0.6

0.8

1

1.2

0 0.1 0.2 0.3 0.4 0.5

Edif

Final MF's on Edif

0

0.2

0.4

0.6

0.8

1

1.2

0 0.1 0.2 0.3 0.4 0.5

Edif

Figure 8.17. Initial andfinal membership functions onEdif: 232-partition
The classification results of the training data are illustrated in Table 8.9.
The classification results indicated superior classification rates with ANFIS com-
pared to the fuzzy clustering and the backpropagation neural network. A total
of 5 trash objects out of 213 trash objects were misclassified resulting in a clas-
sification rate of 97.653%.

Table 8.9. Classification results of trash objects

in the training data using ANFIS: 232-partition

Total number Actual Classified trash type

of objects trash type Bark Stick Leaf Pepper

18 Bark 16 2 0 0

34 Stick 0 33 1 0

54 Leaf, 0 1 53 0

107 Pepper 0 0 1 106