This course is an exploration of an emergent field in computer science, evolutionary computation. EC is a problem-solving technique that uses computer simulations of natural evolutionary processes (as opposed to traditional ‘computer programming’) to model complex behavior and produce fully working ‘undesigned objects’ (which can even include computer programs!). We will examine the three traditional variants of EC, namely, evolution strategies (ES), genetic algorithms (GAs), and genetic programming (GP). We'll look at the theory behind EC, some applications of EC, and current research on EC.
You are expected to bring whichever textbook or handout we're currently working from, your notebook, a writing implement, and a backup writing implement to every class meeting.
[ Back to the Table of Contents ]
In brief, the graded components of this course are three 50-minute exams, research exercises using Learning Classifier Systems, a presentation to the class on current EC work, and participation. Of course, you are also expected to attend each class and lab session, keep up with the reading (see the course schedule), and attend office hours or make an appointment to see me if you are having difficulty understanding the material.
| component | weight | when | ||
|---|---|---|---|---|
| 1. | Examinations |
one | 20% | due February 6 at 10:30 a.m. |
| two | 20% | due March 8 at 3:30 p.m. | ||
| three | 20% | due May 8 at 2:00 p.m. | ||
| 2. | LCS exercises | 20% | April 3–May 3 | |
| 3. | Class Presentation | 10% | May 8, 2:00–5:00 p.m. | |
| 4. | Participation | 10% | throughout the semester | |
Each component of your grade will be assigned a number between 0 and 100, inclusive. Your final grade in the course will be a weighted sum Σ of all components, where the weights are given above. Σ will be converted into a letter grade according to the following scale:
|
|
|
|
|
The weighted sum Σ indicates your minimum grade for the course. Significant and constant improvement throughout the semester may be taken into account in determining the final grade.
[ Back to the Table of Contents ]
The exams are closed-book and closed-notes. They will each be designed to take 50 minutes to complete (though you will be allowed 75 minutes). You will be given each exam in a sealed envelope on the date we review the material for that exam. Each exam is due on the date and time noted on the course schedule.
The first exam will cover the classical theory of natural selection as articulated by Charles Darwin. The second exam will cover the general concept of evolutionary computing and the major variants of evolutionary algorithms. The final exam will cover Learning Classifier Systems. It may also have a small cumulative component.
In the latter part of the semester (from approximately April 3 on), we'll be focusing on a particular branch of EC research, Learning Classifier Systems. We'll be reading LCS papers beyond the textbook, and will use software to explore LCS. In association with this, you'll be assigned some short homework assignments and a longer project or two to work on.
You will make a short presentation to the class on the date the registrar assigns for our final exam. Your presentation will be about a current paper on EC research presented at the most recent Genetic and Evolutionary Computation Conference, GECCO 2006.
The presentation will take place in the time slot the registrar assigns for our final exam. Attendance at this presentation is mandatory. Keep this in mind as you formulate your post-semester travel plans.
Details about the presentation will be provided shortly after the spring break.
You can participate by taking part in class discussions, by asking questions that indicate you have read the assigned material, and by intelligently answering questions I may ask you during class. I expect you to do all three of these things throughout the semester.
Occasionally throughout the semester you will be asked to make a brief presentation to the class; your performance will be factored into the participation component of your course grade.
If you believe that you have a valid scheduling conflict with a graded component of this course (e.g., observation of a religious holiday), you must see me before the graded component is scheduled. I will decide whether the conflict is a valid one, and if so, how your graded event will be rescheduled. Note that taking a non-classwork-related trip will not count as a valid conflict. The dates of all exams are clearly indicated on this syllabus, so look them over carefully and mark the dates on your calendar.
[ Back to the Table of Contents ]
You are expected to be familiar with the Hamilton College Honor Code. In short, any work you do for a graded component of this course must be your own.
Your academic conduct in this class is largely a matter of common sense. You are encouraged to discuss computer science topics covered in class with your fellow students. You are not encouraged to steal their work or to do their work for them.
[ Back to the Table of Contents ]
The Department of Computer Science provides laboratory space, computer equipment, and software for your use in this course. You may only use the hardware and software that you have been authorized to use. We expect you to treat all equipment with the utmost respect and care. Modifying the configuration of any equipment without authorization is prohibited. Please report problems with labs or equipment to our department director of laboratories, Nick Brockner. His extension is 4289.
You will be given an account on my research computer, holland, for your work with Learning Classifier Systems. You should use your holland account only for work related to this class. Your holland account will expire shortly after the end of the semester.
[ Back to the Table of Contents ]
I encourage you to drop by during my office hours to discuss any aspects of the course or anything else you want to talk about. If you can't make my office hours, feel free to make an appointment. The best way to make an appointment is to talk to me after class. The next best option is to contact me by email. You can also call my office (the extension is 4816), but unless you call during office hours, I probably won't answer, and I'm not very reliable about checking my voice mail.
[ Back to the Table of Contents ]
If you have a documented disability and require accommodations to obtain equal access in this course, please contact me at the beginning of the semester and when given any assignment for which accommodation is required. Students with disabilities must verify their eligibility by contacting the Associate Dean of Students for Diversity and Accessibility, Allen Harrison (K-J 104; telephone extension 4021).
[ Back to the Table of Contents ]
[ Return to the CPSCI 307-01 homepage ]
Brian J. Rosmaita <contact me>