06-08775 Introduction to AI

Locations

Recommended Books

Detailed Syllabus and Relevant Links

Week	Lectures	Worksheets	Further Info
1	What is AI? html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf	A link to Marvin Minsky's BBC interview (mp3 file, ~8.5 MB) can be found here as http://web.media.mit.edu/~minsky/BBC3.mp3.
2	Expert Systems html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf	Installation instructions for Worksheet 2 are at INSTALL-expert-system.
3	Decision Tree Learning html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf	Installation instructions for Worksheet 3 are at INSTALL-decision-tree.
4	Neural Nets, Naive Bayes html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf	Installation instructions for Worksheet 4 are at INSTALL-naive-bayes.
5	Genetic Algorithms html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf	Installation instructions for Worksheet 5 are at INSTALL-genetic-algorithm.
6	Knowledge Representation I (Data vs knowledge, Frames, Semantic Nets, Inheritance) html, text only, ascii, gzipped PostScript, pdf.	Assignment available as html, txt, gzipped PostScript, pdf Deadline Friday, 10 December 2004, 6pm (extended from Thursday, 9 December 2004, 12:00 noon)
7	Knowledge Representation II (NL, Translation to Logic, Defaults, non-monotonic reasoning) html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf	Some information on how to use the shell can be found here. (sxi format)
8	Search (uninformed, informed, A-star) html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf
9	Planning (Strips, GraphPlan) html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf
10	Agents - Reactive vs Deliberative html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf
11	Philosophical Issues html, text only, ascii, gzipped PostScript, pdf.	html, txt, gzipped PostScript, pdf
12	Revision	html, txt, gzipped PostScript, pdf

Worksheet 12 should give an idea of the examination paper and the resit paper (Note that past exam papers are not related to the current one.). The exam will consist of FIVE Questions of which you have to choose FOUR. Each question is in style similar to the questions in Worksheet 12. That is,

• it is mainly (but not necessarily exclusively) related to one of the main topics of the lecture (as taught in weeks 2 through 9);
• it is worth 25 marks (out of the 100);
• it consists of four parts, (a), (b), (c), and (d)-parts. The (a)-parts are worth 7% each, the (b), (c), and (d)-parts 6% each.
  The (a) and (b)-parts are supposed to be easy and should be easily solved by anyone who seriously worked on the module. They provide together a third class mark (up to 52% and are sufficient to pass the exam).
  The (c)-parts are more advanced and require to apply known (non-trivial) techniques to unseen examples. With the (a), (b), and (c)-parts together students can achieve 2nd class results (in total up to 76%. While it's theoretically possible to achive a first class this way, practically it's hard not to lose any points, so realistically students who attempt only (a), (b), and (c)-parts will get at best an upper second).
  The (d)-parts require deep understanding of the material and application to problems which students find typically difficult. They are required in addition for a 1st class mark.

Errata to the lectures and exercise sheets can be found from html, txt, gzipped PostScript, pdf

For instructions on how to install the Intro to AI code on your home computers, look at INSTALL-at-home.txt.

Please register as a group here.

Information about the assessment

The module is to 30% continuously assessed. Assignment available as
html, txt, gzipped PostScript, pdf
Deadline Friday, 10 December 2004, 6pm (extended from Thursday, 9 December 2004, 12:00 noon). A test harness is available. Installation instructions are as for all previous source code releases and can be found as assignment.tar.gz in /bham/common/courses/ai-intro/source. We strongly suggest you read the code in spam-filter.scm.

Information on the continuous assessment is available locally from here.

70% of the assessment is by examination in spring 2005. The combined provisional marks for continuous assessment and exam are available locally here.

Some hints about learning

• Take notes, summarise or highlight the content of the lectures, of your reading.
• Revise lectures within 24 hours (in order to maximise the learning effect).
• Use a folder in which you keep all the material for this module like annotated handouts, worksheets, rough notes, returned exercise sheets. Bring this folder to the exercise classes.
• You are expected to spend 1/6 of your working time on this module, that is, at least 6 1/2 hours per week.
• Try to solve the exercises. They are much more important than rote learning since they show whether you understand the material.
• You are responsible for your learning. Act responsibly: Don't be lazy. Enjoy the module; it will be enjoyable only if you work hard.
• In general be active in learning.

Maintained by: Manfred Kerber, School of Computer Science, The University of Birmingham
Last update: 6.6.2005.
The URL of this page is   http://www.cs.bham.ac.uk/~mmk/Teaching/AI/index.html.