Talk by Dr. Lee Spector, Hampshire College
Computational systems based on principles of biological evolution can serve as powerful problem-solving systems while also providing useful models of evolutionary processes. Systems in which computer programs themselves evolve are of particular interest, because a vast array of problems can be phrased in such a way that their answers will take the form of computer programs (or sometimes mathematical formulae, which can be considered to be specific types of computer programs). Any such problem might conceivably be solved by a program evolution system. Furthermore, a vast array of biological processes, including the process of evolution itself, can be modeled by computer programs, and such models might also be suitable targets for program evolution systems. In this talk I will show how the use of expressive representations and insights from biology, specifically concerning the nature of variation and selection, can enhance the power and utility of program evolution systems. I will discuss applications ranging from software engineering to bioinformatics, and ways in which the described systems might help us to better understand fundamental principles of evolution.
Events are free and open to the public unless otherwise noted.