A nonlinear classifier using an evolution of Cellular Automata
Jetsada Ponkaew, Sartra Wongthanavasu, Chidchanok Lursinsap · 2011
Generalized Multiple Attractor Cellular Automata (GMACA) is a special class of Cellular Automata (CA) for nonlinear pattern classification however the disadvantages of GMACA are: there is only one rule vector for classification, a search space for constructing an appropriate graph is exponential growth, and the complexity of classification is O(n2). For this reason, this paper proposed Two-Class Classifier Generalized Multiple Attractor Cellular Automata with artificial point (2C2-GMACA+). It utilizes two-class classifier architecture basis that enables to process two classes at a time. Moreover, exploring an appropriate pivotal point (artificial point) is offered in order to reduce the complexity of classification and search space. The experiments on error correcting capability show that the performance of classification on 2C2-GMACA+is more superior to GMACA.