CSCI 455 Syllabus

Spring 2008


Description

Prerequisite: CSCI 355
A survey of artificial intelligence concepts, theory, and practice. Topics include AI languages, knowledge representation, search strategies, logical and probabilistic reasoning, machine learning, natural language processing, expert systems, computer vision, and AI robotics. Students will implement intelligent systems in software and/or hardware.

Major Topics


Objectives


Course Outcomes

Upon successful completion of this course, a student will be able to


Text

Probabilistic Robotics by Thrun, Burgard and Fox, MIT Press, (ISBN 0-262-20162-3).

Papers and materials from a variety of other sources will also be used.


Homework

All homework assignments are due at the beginning of class. Homework will typically emphasize theory over programming, but may include both.

Format

Each problem you submit will include a description of the problem along with the solution. You must show all work to receive full credit.

Deadline

Homework is due to me at the beginning of class. Late homework will not be graded and will receive a score of zero unless you have a good reason (as judged by the instructor).  If you are sick, contact me before class to make arrangements.

Homework Discussions on WebCT

You are encouraged to discuss the homework on WebCT, and read any posted news before beginning and submitting any homework assignment.

Working Together

Feel free to work in groups, but do not simply copy homework answers from others. Copying work from current or previous CSCI 455 students is considered cheating even if you are only copying a small portion.


Quizzes

Quizzes may be given from time to time and will cover the assigned reading and/or topics from previous lectures.


Labs

There will be four to six labs assigned during the course. Please check the class schedule for details.

The labs will emphasize building and programming robots more than theory. We will be using Eclipse and Lejos (Java for NXT) on Windows XP to program robots. Both tools also run on Linux and Mac OSX if you choose to use either of those platforms.

Groups

You may work in teams of two students on each lab. I suggest that you do labs as a team, but please be equally yoked. Each student in the group should contribute to each lab. In your lab submissions, I require you to give a breakdown of the time each team member spent on the lab; if the hours are greatly out of balance, I reserve the right to change one or more partner's grades.  Make sure that you both work equally hard. Additionally, please share ideas with other teams, but do not be unethical and submit two copies of a program developed by two or more teams or copy code segments from another team.

Submitting Labs

Most labs will be passed off in person to the Instructor. "Pass off" means you will demonstrate the program or robot to the instructor. You should expect to make some fixes or corrections before submitting a final version of a lab.

ALL lab submissions should have two parts:

Failure to follow these instructions will result in receiving a zero for the lab.

Late Days

You may turning any lab up to 1 week late for a 10-point penalty. After 1 week the lab will be given a score of 0.


Exams

Two midterm exams and one final exam will be given. Check the class schedule for dates. The midterm exams will be take-home exams. A calculator is a must have item. The final exam will be given in class. The final exam will not be given early; please don't ask me to disregard college policy.

The topics schedule and topics for the two midterms can be found on the class schedule. The final will be comprehensive, but will emphasize topics covered later in the semester.


Grading

Assignments and exams given throughout the semester are categorized and weighted according to the following schedule. The weighting shown here is approximate, and may be adjusted by the instructor:

Category

Weight

Homework and Quizzes

10%

Labs

28% (7% each)

Final Project

16%

Midterm I

13%

Midterm II

13%

Final Exam

20%

Self-Evaluation

Failure to submit this will result in a one-grade penalty

Grading Scale

Grades will be adjusted slightly at the end of the semester to account for shortcomings in my effort to relay concepts to you. Such adjustments made to the grading scale will only help you get a better grade. The percentages in the following chart show the grading scale:

Grade

Percent

A

100% - 90%

B

89% - 80%

C

79% - 70%

D

59% - 60%

F

Below 60%

Class Participation

It is important for everyone to take an active role in class discussions, questions and answers. Be willing to both answer and ask questions, draw answers on the board. Answer with confidence, even if you think your answer may be incorrect. Additional ways to receive credit are to draw solutions on the board, help the instructor learn something new, help find and correct errors to the website, make real suggestions on how to improve the labs and the course. Participation credit is around 0.5%.

Self-Evaluation

Your ability to evaluate your own growth and work accurately is important. Turn in an accurate self-evaluation, at the end of the term, that includes the following items:

  1. State the grade you think you deserve (stating that you think you deserve a particular grade does not imply that you will receive that grade if your work does not warrant it).
  2. Support and defend your statement accurately, citing projects, effort, growth, scores and any other significant factors.

Evaluations should be no more than 1 page single-spaced. The quality of your evaluation may have an impact on your final grade.
 

Online Grades

Your grades are available online WebCT.  Your instructor will do his best to get things recorded correctly and promptly, but if something is not correct, it is your responsibility to point this out to your instructor.  Check scores regularly.


Cheating, Harassment, and Other Ethical Considerations

You may share ideas on homework and labs, but dividing up work and sharing answers, or submitting another person's work as your own is unethical. You must do your own work.  Violations of the honor code will result in an F for the course, and possible addtional action according to Citadel policies.