The Robotics Institute

Carnegie Mellon Robotics Institute

In my work in genetic algorithms, I have isolated the beneficial effects of commonality-based selection. My current focus is to demonstrate that this feature can supplant recombination as the primary advantage of crossover (sexual reproduction). For example, it can be transferred to "single-parent" operators by using commonality-based mutation--changes are allowed to only "uncommon" components. To be used in hill climbing, simulated annealing, and tabu search; a reference solution will be required.

Research Interest Keywords

genetic algorithms, machine learning, scheduling

The Robotics Institute

Menu