Course Schedule

[Note: this webpage last modified Thursday, 01-Dec-2011 14:37:33 EST]

We will cover perhaps half of the content in the required textbook. This web-page will be kept up to date with the required reading to perform before each lecture.

The following is a best guess, and gives a tentative outline of the semester, organized by week. This webpage will be updated throughout the semester as things change.

Potential new topics: Using game-players to solve problems.

Week 0 (Aug 25)

Reading (required):

• "Computing Machinery and Intelligence" by Alan Turing, in Mind, Vol. 59, No. 236, pgs. 433-460. Can view the original on campus through JSTOR.
• If you have not programmed in Python before, you should be learning the syntax of Python and writing programs for practice (e.g., make a little game, factor integers, financial calculator, numerical solver, ...). See the syllabus for links on Python.

More reading/sources (not required):

• Chapters 1, 2 of the text.
• "Some Challenges and Grand Challenges for Computational Intelligence" by Edward Feigenbaum, JACM, Vol. 50 Issue 1, pgs. 32-40. PDF
• Other visions of CS, etc. in the 50th anniversary issue of JACM
• CACM, artificial intelligence
• "important" AI publications

Week 1 (Aug 30, Sept 1)

Topic: Search (parts of chapters 3-6). This includes alpha-beta pruning for game-playing, which we may use for a project. We could also implement search algorithms on the robotos for a project.
Review of basic uninformed search (BFS, DFS, iterative deepending), including time/space analysis. Casting many types of problems as search problems (e.g., SAT, various puzzles, route planning, ...). Informed (heuristic) search, heuristic functions, A* search

Reading: Chapter 3, Chapter 4 up through section 4.2

Week 2 (Sept 6, 8)

Topics: Search (parts of chapters 3-6). Local search techniques (hill-climbing, simulated annealing, genetic algorithms). CSP's, search on those.

Reading: Section 4.3. Chapter 5.

Week 3 (Sept 13, 15)

Topics: Search (parts of chapters 3-6). Adversarial search, playing games.

Reading: Chapter 6.

Week 4 (Sept 20, 22)

UPDATE, Topics: Boe Bot on Tuesday, then into the learning part. We will try skipping more or less straight to neural nets. We may need to backtrack a bit, so we'll see how it goes...

Boe Bot information: the Boe Bot robotics student guide is here.

Some documentation is here. Look at the Basic Stamp Syntax and Reference Manual 2.2 for a reference on the language and on the capabilities of the board (how much memory, etc.). More stuff is linked from http://www.parallax.com/go/Boe-Bot. You can also look at wikibook about the language PBASIC. And of course, you can search for more information...

We will be using the BS2 rev J on the Board of Education USB rev D. Important point - we only have 26 bytes for variables, and only 2K bytes for our program!

WARNING, STATIC - before touching the Boe Bot, touch something else that is grounded (e.g., computer case) to discharge static.

Week 5 (Sept 27, 29)

Topics: Neural nets, support-vector machines.

Reading: Sections 20.5, 20.6, 20.7

Supplemental reading: LIBSVM, wikipedia article on support vector machines, MNIST handwritten digits database

Note -- cutoff for first exam. First exam includes everything we did up through this week (including neural nets, and some basics of support-vector machines).

Week 6 (Oct 4, 6)

Topics: Last week continued.

Week 7 (Oct 11, 13)

Topics: Order of topics we'll cover over the next few weeks... Learning decision trees. Basics of probability (needed to fill in some of the gaps about learning). Bayes' networks.

Reading: Chapter 18. Chapter 13 (skip section 13.7). Chapter 14.
another source on Bayes' networks, and another one that includes the variable elimination algorithm

Week 8 (Oct 18, 20)

Topics: continuation...

Week 9 (Oct 25, 27)

Topics: continuation...

First exam on Thursday October 27

Week 10 (Nov 1, 3)

Topics: continuation...

Potential (supplemental?) reading: Large margin classification using the perceptron algorithm by Freund and Schapire. Chapter 4 Perceptron Learning of Neural Networks - A Systematic Introduction by Raul Rojas.

Week 11 (Nov 8, 10)

Topics: continuation...

Week 12 (Nov 15, 17)

Topics: Back to chapter 20 Learning Theory. Also, Game Theory and Mechanism Design.

Reading: Sections 17.6, 17.7 Chapter 20.

Week 13 (Nov 22)

Topics:

Week 14 (Nov 29, Dec 1)

Topics: More neural net stuff. Note that linear regression and logistic regression are very similar to perceptron learning with threshold and sigmoid. Perceptron learning with data transformation, and the kernel trick.

Supplemental sources of information: about perceptrons and the kernel trick, and another

Week 15 (Dec 6, 8)

Topics: What is AI, what is the future of AI, etc.?

Final Exam slot (Dec 13 3:00pm)