Genetic Algorithms in Noisy Environments

摘要: Genetic algorithms are adaptive search techniques which have been used to learn high-performance knowledge structures in reactive environments that provide information the form of payoff. In general, payoff can be viewed as a noisy function structure being evaluated, and learning task an optimization problem environment. Previous studies shown genetic perform effectively presence noise. This work explores detail tradeoffs between amount effort spent on evaluating each number evaluated during given iteration algorithm. Theoretical analysis shows that, some cases, more efficient results from less accurate evaluations. Further evidence is provided by case study obtain good registrations digital images.