Knowledge Test
Now that you have completed the tutorial you should understand the fundamentals of Single Layer Perceptrons and MultiLayer Perceptrons. To yourself on this newely aquired knowledge I have composed some questions which appear below. Go through the questions and write your answers down, then click on the link for the answers page. If you go the answer correct then well done, if not then go back to the relevant section of the tutorial so you can brush up on this weaker area.
Questions:
1. What are the basic characteristics of our
"Neuron Model"?
2. What type of system is our "Neuron
Model"?
3. fh(x) is the
______ function. Just fill in the blank.
4. In the case of our "Neuron Model"
the function in Q3 is what type of function?
5. Draw a annotated sketch of our "Neuron
Model".
6. Perceptrons use what learning paradigm?
7. How does the above learning paradigm teach
the Perceptron?
8. Name the steps in the Perceptron learning
algorithm.
9. What controls the adaption rate?
10. Under what circumstances shall the Perceptron learning
algorithm end?
11. What type of problems can a single layer perceptron
solve? Give an example using 2-d pattern space.
12. Name and draw a function that can be used in overcoming
the credit assignment problem.
13. Draw a annotated sketch of how the multi layer
perceptron unit used by example 4 may look.
14. What is the learning rule used in Multi Layer Perceptrons?
15. Explain how the above rules works.
16. What type of problems can MultiLayer Perceptrons solve? Give
an example which also shows what a single layer perceptron cannot
solve.
Previous:- Example 4: Two Dimensional Pattern Space revisted