Title :
The K-winners-take-all neural network of classification
Author_Institution :
Informational Technol. Dept., East Eur. Nat. Univ. named after Lesia Ukrainka, Lutsk, Ukraine
Abstract :
The K-Winner-Take-All (KWTA) classifier is presented in this article. Its architecture is demonstrated. The basic steps of KWTA classifier are described. Its algorithm is presented. The KWTA structure is demonstrated. Its advantages and limitations are analyzed.
Keywords :
neural nets; pattern classification; KWTA classifier; KWTA structure; classification problem; k-winner-take-all classifier; k-winners-take-all neural network; Artificial neural networks; Biological neural networks; Classification algorithms; Computer architecture; Heuristic algorithms; Hopfield neural networks; Classifier; Dynamical rule; K-Winner-Take-All; Subnet;
Conference_Titel :
Experience of Designing and Application of CAD Systems in Microelectronics (CADSM), 2013 12th International Conference on the
Conference_Location :
Polyana Svalyava
Print_ISBN :
978-1-4673-6461-4
