smithj@cs.sjsu.edu) is usually better for reaching me than the phone (408-924-5153), since I check messages more frequently, and I'm often too busy with a student to answer the phone.
Each of the eight major parts of the text will be covered in this class. Of these, the first (Artifical Intelligence) and the last (Conclusions) will be covered only very briefly in the lectures. Problem Solving (especially Chapters 3 and 4) and Knowledge and Reasoning (especially Chapters 7-9) will be covered the most thoroughly. The coverage of Planning and Learning (primarily from Chapters 11, 18, and 19) will be somewhat less thorough. The coverage of Uncertain Knowledge and Reasoning and Communicating, Perceiving and Acting (primarily from Chapters 13, 14, and 22) will be still less so.
For each exam or assignment, numeric grades are given and intervals for each letter grade are assigned (usually 90% for A-, 80 for B-, etc.). To find your course grade, I compare the sum of your numeric grades to the sum of the intervals, except that I often increase the course grade for those with just one poor grade, or who improved throughout the course. The intervals for + and - grades are rather small. My standards for I grades, for makeup exams, and for extending assignment due dates are quite strict. At a minimum, I expect documentation of why you cannot complete the work in the expected time.
See the separate sheets on Assignments and Documentation for specific requirements for submissions. These, a tentative class calendar, and other useful documents will be available on the class web page at
The most important consequences of this policy are that, unless unless I explicitly specify otherwise, work you turn in for this class should be entirely your own, and you should not share your work with anyone else. Additional information on this topic is available on the class web site.
Upon successful completion of this course, students should be able (by hand
or by implementing an appropriate algorithm in some programming language) to:
find solution nodes in a state space using the A* algorithm
translate sentences in first-order logic to conjunctive normal form (CNF)
find proofs by using resolution
Students should be able to explain the advantages and disadvantages
of the following techniques. In cases where no technique is given
as a standard for comparison, students should be able to supply
one or more names of appropriate techniques.
breadth-first search, compared to depth-first search
informed search, compared to uninformed search.
hill climbing
forward checking in constraint satisfaction
alpha-beta pruning
the STRIPS representation for planning
Students should be able to describe
the frame problem
default reasoning
possible representations for time and for beliefs