2010/05/24 by Yakov Karandashev, Boris Kryzhanovsky, Karandashev, Yakov +3
Computer Science · Physics and Astronomy · #Disordered Systems and Neural Networks (cond-mat.dis-nn) #FOS: Computer and information sciences #FOS: Physical sciences #Matrix Theory and Algorithms #Neural and Evolutionary Computing (cs.NE) #Photonic Crystals and Applications #Theoretical and Computational Physics #cond-mat.dis-nn #cs.NE
paper · pdf · doi:10.48550/arxiv.1005.4285
10 pages, 8 figures
arxiv created 2010/05/24 · openalex publication_date 2010/05/24 · arxiv updated 2010/05/25 · openalex created_date 2022/10/03 · openalex updated_date 2026/07/28
The local minima of a quadratic functional depending on binary variables are\ndiscussed. An arbitrary connection matrix can be presented in the form of\nquasi-Hebbian expansion where each pattern is supplied with its own individual\nweight. For such matrices statistical physics methods allow one to derive an\nequation describing local minima of the functional. A model where only one\nweight differs from other ones is discussed in details. In this case the\nabove-mention equation can be solved analytically. Obtained results are\nconfirmed by computer simulations.\n