Abstract
Conventional analyses on dispersion tend to concentrate on the derivation and validation of closed-form solution, which is localized, viz. without an effective analysis pattern to uncover the full picture about dispersion space. This work shall provide a systematic treatment towards the panorama of dispersion space with the assistance of neural network. A thorough analysis pattern is presented to fully explore the dispersion space and gain some insight into it on the basis of several characterizations and rational-constrained analytic domain. The full picture of dispersion space and corresponding dispersion-manipulation capability are disclosed in this manner. The analysis pattern is of generality since the steps involved shall be easily transplanted to other periodic structures.
Disclosure statement
No potential conflict of interest.