Paper

Benchmark Learning Algorithm


Authors:
Anshi Xie
Abstract
In this paper, the benchmarking learning algorithm (BLA), was proposed according to the benchmark learning theory in the business management. In BLA, a competitive learning mechanism based on dynamic niche was set up. First, by right of imitation and learning, all the individuals within population were able to approach to the high yielding regions in the solution space, and seek out the optimal solutions quickly. What is more, the premature convergence problem was solved through new optimal solution policy. Last but not least, BLA is able to accurately detect the slight changes of the environments and track the trajectory of the extreme points in the search space. And thus, it is naturally adaptable for the dynamic optimization problems. In this paper, the main differences between BLA and the existing intelligent optimization methods, such as genetic algorithm (GA), particle swarm optimization (PSO) et al were analyzed and revealed. The comparative experiments for both the static optimization problem and the dynamic optimization problem showed that BLA is robust and able to perform friendly interactive learning with the environments, whose search speed, optimization ability and dynamic tracking ability were far superior to other similar methods.
Keywords
Benchmark Learning; Search Pattern; Evolutionary Algorithm; Swarm Intelligence; Dynamic Environments
StartPage
68
EndPage
79
Doi
Download | Back to Issue| Archive