COSC 410: Artificial Intelligence

CPMA 580: AI/Cognitive Science

Spring 2014

Dr. Donald Simon
Office / Phone: 416 College Hall / 396-6472
Office Hours: MW 5:00 PM - 6:00 PM or by Appointment
E-mail: simon@mathcs.duq.edu, or simond@duq.edu, but not simon@duq.edu
Home Page: www.mathcs.duq.edu/profs/simon.html
Class Page: www.mathcs.duq.edu/simon/Spring14/cs410.html

Text: Artificial Intelligence: A Modern Approach, Third Edition, by Stuart Russell and Peter Norvig

Text Web-site: http://aima.cs.berkeley.edu

Course Objectives: This course is a survey of the area of artificial intelligence (AI). We will discuss what AI is, how it differs from traditional computer science. We will focus on various domains of AI, specifically, natural language processing, automated reasoning, vision, expert systems, and automated learning. In addition, we will study the common sub-problems of these domains, such as knowledge acquisition and representation and search. The course will also cover neural nets. As part of the work in the course, students should expect to become proficient in the Prolog language.

Assignments and programs are due at class time. Tests are in-class and closed book.

Grading Scale:

100-90 = A, 89-80 = B, 79-70 = C, 69-60 = D, below 60 = F.

Plus/minus grading will not be used.

Honor Policy: Students in this class fall under the mandate of the University's academic integrity policy. Any student guilty of plagiarism will receive a grade of ``F'' for the course and will be reported to the College's Academic Integrity Committee. Work done in this course is to be by the individual, not a group. You may not share (copy, give, show) your homework with other students in the course. Any code not your own that is included in your programs must be properly cited. This includes code from the book and that given by the professor. Submitted work must be your own work, although you may include code from the book (or book's web site) and that given out by the instruction. If the code is not your own, it must be properly attributed.

Project:A project addressing a problem in artificial intelligence is required. The project should be the equivalent of three weeks of homework in terms of effort. The topic is of the student's own choosing, with the consent of the professor. The project should involve programming, mathematical analysis, and/or statistical analysis in a significant way. If the program contains code, source code must be included with the project. A write-up of 5-10 pages is also required.

Examples of projects are: An expert system for tree identification, a spam filter, a chess program, a sudoko solver, a simple, domain-specific question and answering system, a neural net to solve a classification problem. The project could solve a problem in a new way, or could be an implementation of a known approach to a new problem.

Late Work: A homework or program assignment will lose 5 points per day that it is late. Homework assignments and programs may not be turned in after they are discussed in class. Weekends are counted as one day. Programs and homeworks are due at class time. All programs and homework must be turned in by 4/23/14. Work turned in after that time will not be accepted.

Students with Disabilities: Students with documented disabilities are entitled to reasonable accommodations if needed. If you need accommodations, please contact the Office of Freshman Development and Special Student Services in 309 Duquesne Union (412-396-6657) as soon as possible. Accommodations will not be granted retrospectively.

Additional Information for CPMA 580 Homework asssignments will include the homework assignments for COSC 410 but will have additional questions for those enrolled in CPMA 580. The mid-term and final will also have additional questions for those in CPMA 580. The project will require a component that involves area of computational mathematics, e.g., Markov Chain Monte Carlo simulations, statistical inference, mathematical analysis.

	Date	Topic(s)	Readings	Slides
1.	1/8	Introduction; Search	Chap. 1-3
2.	1/15	Informed Search	Chap. 4,5
3.	1/22	CSP, Logical Agents	Chap. 6,7	Slides 1
4.	1/29	First-Order Logic	Chap. 8	Slides 2 Slides 3 Slides 4
5.	2/5	Automated Reasoning	Chap. 9, Prover9 home page, Prover9 manual Prover9 examples, Prolog Manual
6.	2/12	Planning	Chap. 10,11
7.	2/19	Knowledge Representation	Chap. 12	Slides 5
8.	2/26	Midterm; Uncertain Knowledge and Reasoning	Chap. 13,14
	3/3	Spring Break
9.	3/12	Learning from Examples	Chap. 18
10.	3/19	Knowledge in Learning, Reinforcement	Chap. 19, 21	nn1 slides nn2 slides
11.	3/26	Natural Language Processing	Chap. 22	Lang1 Slides Lang2 Slides Lang3 Slides
12.	4/2	More NLP	Chap. 23	Speech Slides
13.	4/9	Perception	Chap. 24	Vision Slides
14.	4/16	Project Presentations
	4/23	No class - Monday schedule
	4/30	Final - 6:00 pm to 8:00 pm