Read pages 1-8 of Chapter 1 of the book "Empirical Methods for Artificial Intelligence" by Paul Cohen, pages 1-8 (available at http://people.cs.ubc.ca/~hoos/tmp/c1.pdf and not to be further distributed by any means), and answer the following questions as clearly and concisely as possible:
1. Cohen talks about six components common to scenarios in which AI systems are studied. List their equivalents for the case where an arbitrary algorithm is studied. Briefly explain where your components differ from Cohen's.
2. Illustrate the progression from description to prediction to causal explanation using a simple, hypothetical example of an algorithm that heuristically finds shortest paths in a graph. Your example does not have to be realistic, it just needs to illustrate the right type of statements about the algorithm.
3. What is exploratory data analysis and why is it important?
Now, read pages 358-366 of Chapter 9 from Cohen's book (available at http://people.cs.ubc.ca/~hoos/tmp/c9.pdf and not to be further distributed by any means). Don't worry about things you don't understand (such as references to parts of the books you have not read). Answer the following questions as clearly and concisely as possible:
4. Briefly describe the problem of representative samples. When empirically studying the performance of an algorithm, what do we sample?
5. What is the key idea underlying the example of an automated algorithm design technique Cohen mentions briefly?