Title :
A perception algorithm for discerning non-linearity
Author :
Guo-liang, Li ; Xin-xia, Wang ; Chang-An, Wu
Author_Institution :
Center of Network Inf., Xinyang Normal Univ., Xinyang, China
Abstract :
This paper puts forward a new improved perception algorithm based on solution region by analyzing the pattern´s solution geometric characteristic in space. The algorithm could find a better solution vector in solution region on condition that the pattern´s linearity is separable. Otherwise, it indicates the linearity isn´t separable. That is, it can discern that the solution region doesn´t exist. Consequently, this algorithm can discern quickly whether the pattern´s linearity is separable in its solution region and solve its convergence effectively, and finally enhance the speed of calculation.
Keywords :
geometry; pattern recognition; geometric characteristic; pattern linearity; perception algorithm; solution region; Educational institutions; convergence; pattern; the perception algorithm; the solution region;
Conference_Titel :
Bio-Inspired Computing: Theories and Applications (BIC-TA), 2010 IEEE Fifth International Conference on
Conference_Location :
Changsha
Print_ISBN :
978-1-4244-6437-1
DOI :
10.1109/BICTA.2010.5645224