Neurodynamics and attractors in quantum associative memories

Article type: Research Article

Authors: Rigatos, G.G.^{a; *} | Tzafestas, S.G.^b

Affiliations: [a] Unit of Industrial Automation, Industrial Systems Institute, 26504 Rion Patras, Greece. E-mail: [email protected] | [b] Intelligent Robotics and Automation Lab., National Technical University of Athens, 15773 Zografou, Athens, Greece. E-mail: [email protected]

Correspondence: [*] Corresponding author.

Abstract: Quantum associative memories are connectionist structures that demonstrate the particle-wave nature of information and are compatible with quantum mechanics postulates. Following the solution of Schrödinger's diffusion equation, and using the Hopfield memory model, quantum associative memories are developed. It is proved that the weight matrix of quantum associative memories can be decomposed in a superposition of matrices, thus resulting in an exponential increase of the number of attractors (memory patterns). The storage and recall of patterns in quantum associative memories is studied through a numerical example and simulation tests.

DOI: 10.3233/ICA-2007-14303

Journal: Integrated Computer-Aided Engineering, vol. 14, no. 3, pp. 225-242, 2007