Statistical mechanics for neural networks with continuous-time dynamics

• Verfasst von: Kühn, Reimer [VerfasserIn]
• Verfasst von: Bös, Siegfried [VerfasserIn]
• Titel: Statistical mechanics for neural networks with continuous-time dynamics
• Verf.angabe: R. Kuhn, S. Bos
• Jahr: 1993
• Umfang: 27 S.
• Fussnoten: Falsche Namensform in der Verfasserangabe. Richtig: Kühn, Reimer und Bös, S ; Gesehen am 15.06.2023
• Titel Quelle: Enthalten in: Journal of physics. A, Mathematical and general
• Ort Quelle: Bristol : IOP Publishing, 1975
• Jahr Quelle: 1993
• Band/Heft Quelle: 26(1993), 4, Seite 831-857
• ISSN Quelle: 1361-6447
• DOI: doi:10.1088/0305-4470/26/4/012
• Datenträger: Online-Ressource
• Sprache: eng
• K10plus-PPN: 1849778876

Abstract

The authors summarize their current knowledge about the long-time behaviour of networks of graded response neurons with continuous-time dynamics. They demonstrate the workings of their previously developed statistical-mechanical approach to continuous-time dynamics by applying it to networks with various forms of synaptic organization (learning rules), and neural composition (neuron-types as encoded in gain functions), as well as to networks varying with respect to the ensemble of stored data (unbiased and low-activity patterns). They present phase diagrams and compute distributions of local fields for a variety of examples. Local field distributions are found to deviate from the Gaussian form obtained for stochastic neurons in the context of the replica approach. A solution to the low firing rates problems within the framework of nets of analogue neurons is also briefly discussed. Finally, the statistical-mechanical approach to the analysis of continuous-time dynamics is extended to include effects of fast stochastic noise.