Individualized Courses

It is possible to tailor a course to your particular needs as a student. Within CS there are several different course numbers and titles for these individualized courses including CS 3960 Honors Reading, CS 3980 Honors Research, CS 3990 Independent Study, CS 4910 Senior Reading and Research, CS 4960 Directed Readings, CS 4990 Independent Study, CS 5960 Directed Readings, CS 5990 Independent Studies, ECE 5990 Special Studies, CS 6960 Directed Readings, and CS 6990 Independent Study. While these courses differ in some of their details, they share enough similarities that it makes sense to describe them in a single document.

Course Title:

Varies (see above)

Instructor:

Dean Hougen, DEH 242, 405-325-3150, hougen@ou.edu

Class Hours:

Meetings, generally weekly, in my office.

Office Hours:

Additional meetings may be scheduled as needed.

General Information:

If you wish to take an individualized course from me, you will need to create a proposal for the course you want to take. You are encouraged to consult with me during the creation of this proposal and may revise your proposal based on my feedback. If I accept your proposal for a course and have time in my current schedule, I will allow you to register for an appropriate individualized course with me.

Your proposal must include the following elements:

Topic

Your topic must mesh with my research interests. The general topic must be artificial intelligence (AI). Within that topic, I am primarily interested in having students study machine learning and/or robotics but I will entertain proposals in other areas of AI if you can make a compelling case - one that demonstrates your existing knowledge of the area and sparks my interest.

To learn more about my research interests, see the following:

My TEDxOU 2015 talk on the evolution of nurturing and learning in robots

My research statement
My research on robotic intelligence and machine learning
My research on distributed robotics
My research on learning and robotics
My resume (paying particular attention to recent publications)

PLEASE NOTE: While your topic must be consistent with my research interests and I am willing to work with you to help you find an appropriate topic, I will NOT be assigning a topic to you. You MUST arrive at YOUR OWN topic. There are two reasons for this:

Being able to identify research topics that are important, timely, and relevant is an important skill for you to learn and one that will only come with (possibly guided) experience. This is your chance for that experience (with guidance).
If you try but cannot arrive at your own topic, that is a good sign that you don't really have a sufficient interest or background in AI to be taking an individualized course with me. If you are genuinely interested in AI but lack the background, then you should work to get that background (by, for example, taking courses such as Artificial Intelligence, Intro to Intelligent Robotics, etc.) BEFORE asking to take an individualized course with me.

Sources of Knowledge

Because of the nature of individualized courses, I will not be presenting the material to you - you will be accessing sources of knowledge besides your instructor and on your own. You will work to understand these materials on your own and then I will work with you during our meetings to help you clear up questions that you may have over the material.

These sources of knowledge may include textbooks but for more advanced individualized studies most of what you are learning from must be be primary source materials from the peer-reviewed literature (such as conference papers and journal articles).

The following textbooks are good starting places:

Artificial Intelligence: A Modern Approach (third edition), Stuart Russell and Peter Norvig, 2009, Prentice Hall. (ISBN 0-13-604259-7) This is a very good textbook on AI that is used for CS 4013/5013 Artificial Intelligence. If you wish to use it as a source of knowledge for an individualized course, you should explore sections or chapters not covered in CS 4013/5013. Alternately or in addition, you could use its references to move to original source material.
Computational Intelligence: An Introduction, (second edition), Andries P. Engelbrecht, 2007, Wiley. (ISBN: 978-0-470-03561-0) This is a good textbook on artificial neural networks (ANNs), evolutionary computation (EC), swarm intelligence, artificial immune systems, and fuzzy systems. The chapters on ANNs and EC are used in CS 5970 - Artificial Neural Networks and Evolution (ANNE). If you wish to use it as a source of knowledge for an individualized course, you should explore sections or chapters not covered in ANNE. Alternately or in addition, you could use its references to move to original source material.
Evolutionary Computation: A Unified Approach, Kenneth A. De Jong, 2006, MIT Press. (ISBN-10: 0-262-04194-4; ISBN-13: 978-0-262-04194-2. Choice Outstanding Academic Title, 2006.) This book is a good second book on evolutionary computation but should be preceded by a more basic introduction to the topic.
Introduction to AI Robotics, Robin Murphy, 2000, MIT Press. (ISBN 0-262-13383-0) This is a very good textbook in terms of the material covered. Unfortunately, there are quite a few errors in the textbook. For example, there are several places that the pseudo-code presented does not match the text description of it. To get an idea of these types of errors, see Homework 2 from my Spring 2002 Introduction to Intelligent Robotics course and its solution. This book is used as a textbook for CS 4023/5023 - Introduction to Intelligent Robotics. If you wish to use it as a source of knowledge for an individualized course, you should explore sections or chapters not covered in CS 4023/5023. Alternately or in addition, you could use its references to move to original source material.
Behavior-Based Robotics, Ronald C. Arkin, 1998, MIT Press. (ISBN 0-262-01165-4) This textbook is a little more advanced than the previous but should not be difficult to follow.
Machine Learning, Tom M. Mitchell, 1997, McGraw Hill. (ISBN 0-07-042807-7) An excellent text covering a range of machine learning methods as well as some overall theory. While there are few errors, the author has nonetheless thoughtfully provided errata pages which are available on his site for the book.

Methods of Demonstrating Understanding

In order for me to grade your work in this course, you will need to demonstrate to me your understanding of the course subject matter. You will need to demonstrate this understanding in at least two ways:

Demonstration of Understanding of Individual Sources of Knowledge. You will need to demonstrate that you understand the individual sources of knowledge that you are using in this course. This would generally be something along the lines of chapter reviews or article summaries although other methods such as presentations of the material are also possible.
Demonstration of Integration and Application of Understanding of Sources of Knowledge. You will need to demonstrate that you can integrate and apply what you have learned from your knowledge sources. This would typically be something along the lines of coding up one or more of the methods you have learned about, applying this code to a particular problem set of your choosing, and writing up a report of your results. This option would be appropriate for individualized research courses. Other possibilities, such as substantial literature surveys, are also possible and would be appropriate for individualized readings courses. In any case, this will be a substantial piece of work, suitable for conference publication.

Percent of Grade: You will specify how much of your grade should be based on your demonstration of understanding of individual knowledge sources and how much on your demonstration of your integration and application of your understanding of these knowledge sources. This split must be between 70%/30% and 30%/70%.

Schedule: You will need to create a week by week schedule of your activities during the course.

To get started, you can look at a sample syllabus outline with schedule in Portable Document Format or in Open Document Text format.

PLEASE NOTE: I am only willing to consider individualized courses that are at least three credits and run for at least one full semester. I do not believe that you can learn enough about the topic or the process of doing research and publishing results during smaller courses to make the effort (yours or mine) worthwhile.