Artificial Intelligence

EXAM WEEK OFFICE HOURS:   TUES.  MARCH 20, 2012 2-4PM   E2 255

Lectures:

     TTh 12-1:45pm  Thimann Lecture 001

     OPTIONAL labs: (TA will be there)
     BE105  M 10-12, W 2-4

Instructor: Robert Levinson

     Office Hours:  W 4-6PM   in E2 255, after class, or by appointment.

     Phone number: 459-2087
     E-mail: levinson[at]soe[dot]ucsc[dot]edu  - Write me!

Teaching Assistant: John Murray
     Office Hours: (see above labs)
     E-mail:   john[at]lucidbard[dot]com

Reader/Tutor:  ?? (mainly help with projects)

Homework assignments can be found below;

Rules for the current version of RPS-Safari can be found on the left or here.

Lecture notes are posted on the lecture notes subpage.

     Prerequisites:
        CMPS140:

  • Undergraduate standing and CMPS101 or instructor's permission.
  • Knowledge of predicate calculus (first order logic) is helpful, but not mandatory.
  • You should have some previous exposure to recursion.
  • The more computer science you know, the better.


CMPS240:  Available if and only if you are a graduate student.


     Required:
             Artificial Intelligence: A Modern Approach
             Russell and Norvig 2008 Publisher: Prentice Hall
             Third Edition
         ***    call this "AI"   ***
          Is eligible for etxtbook promotion by Amazon for $10 off kindle
              Text is also used in CMPS241.

      Optional - but useful:
         ANSI Common Lisp (paul graham)

      Optional:    (for more practical examples:)
          Practical Common Lisp - Peter Seibel
          *** call this "PCL" ***

         The common lisp Bible and Language standard:
               Common Lisp, Second Edition: The Language
              Guy Steele, Second Edition
           Available on line:  http://www.cs.cmu.edu/Groups/AI/html/cltl/clm/


Also on Line:  A Gentle Introduction to Symbolic Computation
   David Touretsky  http://www.cs.cmu.edu/~dst/LispBook/index.html


(Also see On-Line Lisp Tutorials/Intros on World Wide Web.
For example:
http://www.apl.jhu.edu/~hall/lisp.html
http://www-2.cs.cmu.edu/~ggordon/lisp-hints.txt
http://art2.ph-freiburg.de/Lisp-Course



READ ABOUT RPS  type games at:

http://en.wikipedia.org/wiki/Rock-paper-scissors

and many more places on the web!



Evaluation:

Category Weight # Description
Written Assignments 25% 4 HW1-HW4
Adaptive Programming Project 25% 4 P1-P4
4 Quizzes 25% 4 30 minute exams
Final exam 25% 1 180 minute exam.

YOU WILL BE SCORED  25-25-25-25  weighing all categories equally.

Exceptional performance of your project  agents will be recognized (bonus points).

Minimally, at least 45% must be achieved in each part to be even considered for passing the course.


Policies:
Homework, programming assignments, and project assignments may be done in groups of 1 or  2 or 3 (but CMPS140 students may not partner with CMPS240 students) that remain steady throughout the quarter, and also must be turned in at class time (or other, if specifically  announced).

It is fine and encouraged to discuss homework problems and projects with other  students - BUT CHEATING or ACADEMIC DISHONESTY on any course item (such as direct verbatim copying  from a  member outside  your  group  or  during an exam) will result in not passing the course and other highly undesirable consequences.

    No incompletes for the course will be given.


    Notes:
       About 30 percent of the course material will be about Common Lisp.
       Prior familiarity is not assumed.

       In addition, students will write "Adaptive Pattern-Oriented Game
       Players" in Lisp. The learning methodology used follows
       the instructor's model for experience-based learning and others.

       Tournaments may be run at various points in the quarter and
       up to 40% percent of your project grade will depend on your agent's
       performance.

       Past games invented by the instructor are called "Lowball", "Undercut"
       "Stocker", "Election" and "MasterMind+".
       This year's game is called "RPS Safari".

 

        CMPS 240 STUDENTS WILL BE REQUIRED TO WRITE UP THEIR PROJECTS and their
          agents will be scored against ALL students. The papers will
          include introduction and motivations, strategic approach,
          performance results and analysis.
          Your papers will be shared with the class  at each check point and
         should be workshop paper quality or better.

          CMPS140 STUDENTS will not be scored against  CMPS240 students
         (although their agents will compete in the same tournaments).



          Running Common Lisp:
          sbcl On most cats  machines like unix.ic:
          (quit) to exit  or ^D to exit.

          On soe.ucsc.machines (like sundance.soe.ucsc.edu)  /usr/local/sbcl
          or sbcl should work.

          (we may further specify which Lisp(s) to run when the projects begin.



          Course Equation: Total Information = Diversity + Symmetry.

          Course Affirmation 1: Every day and in every way  my agent and I are
          growing in intelligence and understanding.

          Course Affirmation 2: Exploiting mathematical structure is easy,
                   rewarding and fun.


Homework Assignments and Topics

Submission Instructions


Submission of HW and Projects:

Please submit your homework from here on out in the course homework locker. This can be accomplished on any unix machine by typing:
submit cmps140-ral.w12 <assignment> <file(s)>.

For example, for P2 and HW2, you would submit your files like so:

submit cmps140-ral.w12 P2 agent.lisp monitor.lisp
submit cmps140-ral.w12 HW2 homework2.txt

If you need help getting your homework into txt format. Also note that P2 is distinct from p2. Use the upper case variant.


Besides using a consistent and reliable mechanism for submitting assignments, this also enables you to submit revisions simply by submitting the new version of the file. The latest version submitted before the deadline will be used.

I. Introduction to AI and Lisp Programming

Read Chapters 1,2.1-2.3 in AI.
     and 2,3,5,6,9 in  the Lisp book (or you may learn Lisp some other way)


     HW1=   2.1, 2.3,2.4,2.5,2.6 (5 problems in all - due Thursday, January 26)
     P1 = Problems 2,3,4,5a,7 from:
     http://www.apl.jhu.edu/~hall/AI-Programming/HW1-Review.html
    and Problems 1-4 from:
    http://www.apl.jhu.edu/~hall/AI-Programming/HW2-Higher-Order.html
    and problems 1,3,4 from:
    http://www.cs.northwestern.edu/academics/courses/325/exercises/lisp-exs.html
       (12 problems in all - due Thursday January 26  by Midnight.)


II. Search and Heuristics

Read Chapters 3 and 4.1-4.3 in AI
    HW2: 3.4,3.8,3.9[CANCELLED], 3.14,3.21 ,4.1  (6 problems in all - due Thursday  Feb 9
    P2:  Project Phase II: Game monitors + simple agents  due Thursday  Feb. 9)

III. Knowledge Representation and Logic

     Read  Chapters 7.3-7.6 and 8.1-8.3   and 9.1-9.2 and 9.5 in AI
     HW3:  7.4,7.6, 7.8, 7.10, 8.2,8.3,  8.6, 8.8, 9.3, 9.4
                 (10 problems in all - due Thurs. Mar. 1 at midnight) 5% bonus for getting it in by 6pm. 
     P3: Project Phase III: strategic agents (Due Tues. Mar. 6)

IV. Machine Learning

      Read 19.1, 20.4,20.5 21.1-21.2  in AI
      HW4: 19.1, 19.4[CANCELLED],  18.6,18.11, 18.12   (4 problems)
         +  Learning demo [CANCELLED]

         P4:  Project Phase IV   (due Friday. March 16 @ Midnight)

     : learning agents

Instructors and Assistants