CSC 720

Artificial Intelligence II:
Machine Learning and Adaptive Systems

Spring 2001


Instructor

Instructor
Dr. Dennis Bahler
Phone: 919 515 3369
Fax: 919 515 7896
Email: bahler@ncsu.edu
Office: 216A Withers Hall
Office Hours: Tuesday/Thursday 1430-1600, or by appointment
Teaching Assistant
None

If you have a technical or administrative question that might be of interest to other members of the class, please either ask during class or use the course messageboard. Otherwise, bring your question to the instructor during office hours. I am also highly available via email. Please include CSC 720 in the subject line.

Course Details

Class Time: Tuesday and Thursday 1305-1420
Place: 222 Withers Hall
URL: http://wolfware.ncsu.edu/courses/csc720/
Course Locker: /afs/eos/courses/csc/csc720/

Description

This is a seminar and project course in advanced artificial intelligence, in particular on intelligent computer systems which learn, i.e., adapt their behavior through experience, and system capable of dealing with uncertainty or approximation. The course will focus on readings, discussion, and student presentations interspersed with lectures by the instructor. Readings will come from the technical literature.

Tentative Course Outline

Course topics will be chosen from among: Decision Tree Induction, Neural Network Models, Self-Organizing Systems, Genetic Classifier Systems, Version Space and Candidate Elimination, Inductive Logic Programming, Induction of Bayesian Models from Examples, Unsupervised Concept Formation, Explanation-Based Learning, The Biology of Learning, Learnability Theory and PAC Learning, Learning Automata, Computer-Assisted Instruction, and Automating Knowledge Acquisition.

Objectives

  1. To acquaint students with the current state of research in selected areas of artificial intelligence through reading and discussion of relevant literature. Areas of special interest this semester are machine learning, artificial neural models, and other adaptive systems.
  2. To enable students to develop and synthesize their own and others' ideas relating to some aspect of artificial intelligence (not necessarily one of the areas explicitly covered in class) in the form of either a major paper (20-25 pages) or a significant programming project.

Prerequisite

CSC 520 (Artificial Intelligence I) or the equivalent or permission of instructor.

Grading

Letter grades will be assigned using the plus-minus scale. Grading will be based on a weighted combination of the following elements:
  1. ATTENDANCE, CLASS PARTICIPATION, PRESENTATIONS, etc. (30%): Active participation in class discussion is ESSENTIAL for success in this course. STUDENTS WILL BE EXPECTED TO HAVE READ THE ASSIGNED MATERIAL IN ADVANCE OF CLASS. Each student will be responsible for leading the class from time to time by presenting one or more related papers and initiating the discussion.
  2. SHORT QUIZZES (10%): Students will be expected to have read the material in advance of class. Each presenting student is required to write a short (1 page, 5- to 8-minute) quiz covering the day's reading. All students will take these quizzes, including authors. Grading will be done by the instructor. The author's answers, provided they are adequate, will constitute the answer key for each quiz.
  3. PROJECT and/or PAPER (50%): Either (i) a major paper (20-25 pages) fully annotated with references to the literature as appropriate, or (ii) a project requiring significant programming in some area of artificial intelligence. Projects will require a descriptive writeup and extensive program documentation as well as the code itself. Topics of papers and/or projects must be approved in advance by the instructor. All students will present their projects or papers orally at the end of the course. Time permitting, there may also be progress reports mid-semester.
  4. FINAL EXAM (10%): There will be a comprehensive final exam on the readings at the conclusion of the course.

Text

There is no textbook for this course. Readings will be made available.

Academic Integrity

A student shall be guilty of a violation of academic integrity if he or she represents the work of others as his or her own or aid another's misrepresentation. Any violation associated with an examination or quiz will result in a failing grade for the course. Project related offenses may incur a reduction of up to two letter grades. Such violations will be reported to the Office of Student Conduct, which may impose penalties beyond those by the instructor.

Students with Diabilities

If you have a disability that may affect your participation in this class, please notify the instructor. Students requiring extra time for examinations and quizzes are asked to make arrangements at least three days in advance. You may contact the NCSU Disability Services for Students Center regarding campus services at the Student Health Center for more information and assistance.