Evolutionary algorithms offer the possibility of substantially reducing the programming burden for humans. We will talk about a specific form of evolutionary algorithm, Genetic Programming, and discuss how performance is measured in the field. We will introduce the concept of incremental evolution which has the promise of reducing the computational expense of genetic programming. We will then discuss how incremental evolution might impact standard performance measures, and how it might be automated. Finally, we will discuss an early-termination heuristic that offers a further potential performance benefit to Genetic Programming.

The CVSL seminars are held on Fridays at 11:30 a.m. in room PLT-2744.