Summary

International Symposium on Nonlinear Theory and Its Applications

2015

Session Number:B4L-E

Session:

Number:B4L-E-1

Evolutionary Design of Experiments to Analyze Ultra Large Scale Systems

Takeshi Uchitane,  

pp.716-719

Publication Date:2015/12/1

Online ISSN:2188-5079

DOI:10.34385/proc.47.B4L-E-1

PDF download (366.1KB)

Summary:
Making a design of experiments is very important especially in considering multivariate analyses for ultra large scale systems. It is because that it is too hard to obtain whole or enough analysis data from such systems which have exponentially large parameter space. Thus, a design of experiments which makes it possible to analyze such systems is required. In this paper, a novel framework to dynamically obtain appropriate analysis data sets is proposed. In the framework, an evolutionary algorithm (EA) works so as to optimize analysis results. The results are obtained by analyzing a part of whole data sets. Such sub data set is treated as individuals of EAs. Numerical experiments are performed to make sure that the proposed framework is available to generate appropriate analysis data sets.