An architecture for massive parallelization of the compact genetic algorithm

摘要

This paper presents an architecture which is suitable for a massiveparallelization of the compact genetic algorithm. The resulting scheme hasthree major advantages. First, it has low synchronization costs. Second, it isfault tolerant, and third, it is scalable. The paper argues that the benefits that can be obtained with the proposedapproach is potentially higher than those obtained with traditional parallelgenetic algorithms. In addition, the ideas suggested in the paper may also berelevant towards parallelizing more complex probabilistic model buildinggenetic algorithms.