=============== ABOUT THE FINAL =============== Closed-Book. The final will emphasize the material we covered after the midterm (i.e., from syntax and semantics of 1st-order predicate calculus onward). Note, however, that the material has built upon itself throughout the course. In particular, the final will focus on the lecture material and assignments since the midterm; assignments 4 and 5 and the wumpus project; and the following parts of the text: Focussed material: All of Chapter 6, All of Chapter 7, Chapter 8, pp. 226-231 Chapter 9, pp. 265-285 Chapter 10, pp. 297-298 pp. 304-top 305 p. 310-311, 313 pp. 313-321; 323-325 Chapter 14 Previous material; will not be focussed upon, but may be included as part of a question on the later material: Chapter 2, Chapter 3, Chapter 4 (but not material on SMA* or simulated annealing) The following topics will be focal: syntax and semantics of FOPC (and logical notions such as model, satisfiability, logically follows, etc) be sure to be able to distinguish function from predicate symbols, and know what kinds of things they are assigned under an interpretation. be sure to know what it means to say that a set of logical sentences is consistent, and what it means to say that a sentence logically follows from a set of sentences. know the difference between an implication and a rule of inference know what the closed world assumption is reasoning: unification conversion to clausal form resolution and question-answering representation in first-order predicate calculus, including reification, the situation calculus the frame problem, the qualification problem, frame axioms, closure axioms, preconditions, effects horn clauses, and translating FOPC into Prolog production systems semantic nets and frames, and equivalence to FOPC inheritance (the material that is in chapter 10) strips-style planning systems goal-stack planning non-linear planning Statistical and machine learning techniques: classification graphical and decomposable probability models conditional probability conditional independence assumptions model search (what are the successor, h, and goalp functions?) overfitting the data evaluation: training versus test data N-fold cross validation basic ideas of Bayes' Rule and its use (chapter 14) basics of decision tree induction [There will be no questions about Lisp on the exam] Note: an old final is available to you on the web page. * I look for answers THAT ANSWER THE QUESTION and * that say things that are true. No credit is * given for stating true but irrelevant things. * Please do not just do a "core dump" on the topic of * the question.