Benchmark Electromagnetic Inverse Scattering by Using Differential Evolution - A Big Data Perspective

摘要

The benchmark electromagnetic inverse scattering problem is re-visited in this paper from a big data perspective. It serves as the benchmark application problem in systematic parametric study of differential evolution (DE). Representative strategies with a full sweeping of intrinsic control parameters are applied to draw a systematic picture of DE. Insights extracted from preliminary numerical results are presented to rebut the questionable statements and advise DE applicants.