Prerequisites: CS 782 or permission of the instructor.

This course will serve as an introduction to and comprehensive survey of the rapidly growing field of evolutionary computation. The course will cover in some detail the main paradigms, including genetic algorithms, evolution strategies, and evolutionary programming. These basic evolutionary algorithms will be studied both empirically and theoretically, and their usefulness relative to other more traditional algorithms will be analyzed for a variety of problem domains. In addition, their relationship to the field of Meta-heuristics and the field of Natural Computation will be explored.

Text book: Evolutionary Computation: A Unified Approach (K. De Jong, MIT Press 2006)

There will be regular homework assignments, primarily computational and will involve programming.

A final term project will be required.

Planned list of topics:

1. Evolutionary Computation:
   • general introduction, overview
2. Canonical Evolutionary Algorithms:
   • genetic algorithms (Holland)
   • evolution strategies (Rechenberg)
   • evolutionary programming (Fogel)
3. A Unified View of Evolutionary Algorithms
4. Problem Solving Paradigms
   • survey of applications involving search, optimization and machine learning
5. Formal Models and Comparative Analysis
   • survey of EA theory
6. Recent Developments and Future Directions