Homework 6 - Chapter 8

Each question is worth 2 points.

1. Describe the function of each layer in a neuro-fuzzy system in detail. How does it implement fuzzy inference?
2. How does a neuro-fuzzy system learn?
3. Draw the neuro-fuzzy system that would be trained during the automatic rule discovery process for the following parameters:
   • Input x1 that corresponds to the fuzzy sets A1, A2
   • Input x2 that corresponds to the fuzzy sets B1, B2
   • Output y that corresponds to the sets C1, C2, C3
4. How does a neuro-fuzzy system identify bad rules, either given by an expert or generated by the automatic rule discovery method?
5. Describe each layer in an adaptive neuro-fuzzy inference system.
6. How does the adaptive neuro-fuzzy inference system differ from a neuro-fuzzy system?
7. Describe the hybrid learning method used by an adaptive neuro-fuzzy inference system.
8. Describe how a set of weights are encoded into a chromosome for an evolutionary neural network.
9. Describe the changes made to the genetic operators crossover and mutation for an evolutionary neural network.
10. What is a grid-fuzzy partition? How is it used in fuzzy evolution systems?