Characterization of canonical robust template values for a class of uncoupled CNNs implementing linearly separable Boolean functions
Yih-Lon Lin, Wei-Chih Teng, Jyh-Horng Jeng, Jer‐Guang Hsieh · 2005
Abstract:- In this paper, the geometric margin is used as a robustness indicator of a CNN (cellular neural network) implementing a linearly separable Boolean function. For a class of uncoupled CNNs having low template values, characterization of canonical robust template values is made by finding the maximal margin canonical hyperplane. Support vector machine (SVM) technique is employed for the associated optimization problem. Two illustrative examples are provided to illustrate the main result. Key-Words:- cellular neural network, boolean function, maximal margin 1