CSCI/PHIL 4550/6550 Artificial Intelligence Test 2 Guide Spring 2021 Test 2 will cover material from Chapters 5-6 of the Russell & Norvig text as well as the Gilovich reading. The test will be open book and open notes. Please do not let this lull you into a false sense of security! You will not have time to teach yourself the material during the test! You are also expected to take the test on your own. Per the honor code, do not consult with others while taking the test. The test will consist of a true/false, multiple choice, fill-in-the-blank), and/or short answer questions, and a section of questions requiring longer answers. The latter will typically involve performing calculations/computations by hand (e.g., given a set of clauses, use resolution to prove the clauses inconsistent). Please remember to read through the exam before starting so that you can strategize your approach. The following are among the tasks you might be asked to perform. 1. Given a description of a two-player adversarial game, draw the full MINIMAX game tree and indicate MAX’s best move for a given initial state. 2. Using the same problem, show which parts of the tree are removed by α – β pruning. Note that this does not mean that you need to demonstrate the sequence of steps taken by the pruning algorithm. 3. Draw a game tree for a 2-player game involving chance, again indicating MAX’s best move. 4. Draw a game tree that demonstrates the EXPECTIMINIMAX values for a game of chance. 5. Explain how a heuristic function or forward pruning are used. 6. Describe how well the standard approach to game playing would apply to games such as tennis, pool, and croquet, which take place in a continuous physical state space. 7. Given a constraint-satisfaction problem involving binary constraints and specified domains, draw the corresponding constraint graph and determine whether the graph is arc consistent. 8. Given a binary CSP problem, use AC-3 to make each variable arc-consistent (or to show that the CSP is unsolvable). 9. Given a CSP problem, describe heuristics that could be used to improve search. 10. Given a tree-structured CSP, create a topological sort and explain how this affects the run time and why. 11. Given a CSP, use forward checking to show how it can be made arc-consistent (or is unsolvable). 12. Explain how heuristics can be used to solve CSPs. 13. Describe Gilovich’s account of the representativeness heuristic, describe how it would apply to a mini case study to explain an outcome. Do the same for the clustering illusion. 14. What is statistical regression, and how does it lead to problems in reasoning? 15. How does a tendency to be overly influenced by confirmatory information affect our ability to form beliefs about relationships between two variables (e.g. adoption and conception)? Why do we do this? Explain the card example, and describe how it is related to this problem. 16. What are the reasons why a false sense of confirmation is produced? How does the way in which we search for information affect this? The above list is only a guide and should not be considered complete, and it is not guaranteed that every topic above will make an appearance on the test. As you review the text (including the summary section), your notes, and the slides, you should create additional questions using these as guides to help yourself