Title :
Design of associative memories by using machine learning
Author :
Citko, Wieslaw ; Sienko, Wieslaw
Author_Institution :
Dept. of Electr. Eng., Gdynia Maritime Univ., Gdynia, Poland
Abstract :
Problem of nonlinear mapping-based design of associative memories can be regarded as a covering problem in the case of feedforward architecture and as a generation of fixed points in the case of feedback structure. The machine learning techniques has been used as the analytical tool.
Keywords :
content-addressable storage; learning (artificial intelligence); associative memories design; feedforward architecture; machine learning techniques; nonlinear mapping-based design; Approximation methods; Associative memory; Filters; Hilbert space; Kernel; Large-scale systems; Machine learning; Neural networks; Neurofeedback; Nonlinear equations; design of associative memories; machine learning; nonlinear mappings;
Conference_Titel :
Nonlinear Dynamics and Synchronization, 2009. INDS '09. 2nd International Workshop on
Conference_Location :
Klagenfurt
Print_ISBN :
978-1-4244-3844-0
DOI :
10.1109/INDS.2009.5227991
