A New Parallel Genetic Algorithm

One problem of propagating the globally fittest individual via neighbourhood evolving in both island model and cellular model of existing parallel genetic algorithms (PGA) is that the migration of globally best individual is delayed to non-adjacent processors. That may cause inferior search in those sub-populations. The propagation delay of the globally best individual is proportional to the network distance between two processors. Delayed migration of best individual in parallel genetic algorithms is an essential deviation from sequential version of genetic algorithm, in which the best individuals are always used to compete with other individuals. To solve this problem, this paper proposes an extended version of island parallel genetic algorithm, Virtual Community PGA (VC-PGA). In this paper, VCPGA is applied in a case study of optimizing parameters of back-propagation neural network classifier.

• Type: Conference
• Year: 2002
• Where: ISPAN
• Authors: Ling Tan, David Taniar, Kate A. Smith
• Keywords: Distributed And Parallel Computing, Globally Fittest, ISPAN 2002, Parallel Genetic Algorithm, Virtual Community PGA