IBL Simulation Assignment - CS 572
Using the IBL simulator, do the following experiments. For each experiment give a short (~paragraph) discussion of your findings and observations. Remember. This is a powerful network. You're testing and experimenting with the real thing. Have fun with it.
Note: For all assignments in this course requiring your observations (which is the typical case) the most important thing is not just mentioning what you observe occurring, but I want your explanation as to why it is occurring. If you can't figure it out, then give your best try at explaining it. This is where learning can best occur, when trying to figure out why the models do what they do.
1. Experiment with the IBL algorithms (IB1-IB3) on different training/test sets (MLDB and/or synthetic). Start with simple application and then try more complicated ones. After studying a few applications, discuss the following questions.
a) How long does learning take
b) How accurate is the model
c) How large a concept description is created
d) What is the effect of noise
e) What is the effect of missing attributes
f) What kind of applications are easy/hard for IBL
g) What is the effect of different training set sizes
h) Other observations
2. Use your own "real world" application, test it, and discuss how well IBL handles it.
3. Be creative. Try some of the following and some of your own experiments:
• Experiment with other IBL parameter settings
• Give graphical results of modifying noise, missing values, types of attributes, etc.
• Use "learning rate" as a decision for when to stop learning
• Come up with some of your own research results!
Total pages: ~3-4