Statistical mechanics of lossy compression for non-monotonic multilayer perceptrons

Simple item page
dc.creatorCousseau, Florent
dc.creatorMimura, Kazushi
dc.creatorOmori, Toshiaki
dc.creatorOkada, Masato
dc.date2008-07-25
dc.date.accessioned2026-07-07T09:58:23Z
dc.date.available2026-07-07T09:58:23Z
dc.descriptionA lossy data compression scheme for uniformly biased Boolean messages is investigated via statistical mechanics techniques. We utilize tree-like committee machine (committee tree) and tree-like parity machine (parity tree) whose transfer functions are non-monotonic. The scheme performance at the infinite code length limit is analyzed using the replica method. Both committee and parity treelike networks are shown to saturate the Shannon bound. The AT stability of the Replica Symmetric solution is analyzed, and the tuning of the non-monotonic transfer function is also discussed.
dc.description29 pages, 7 figures
dc.identifierhttps://arxiv.org/abs/0807.4009
dc.identifierhttp://arxiv.org/abs/0807.4009
dc.identifierPhys. Rev. E, 78, 021124 (2008)
dc.identifierdoi:10.1103/PhysRevE.78.021124
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/167697
dc.subjectStatistical Mechanics
dc.subjectDisordered Systems and Neural Networks
dc.subjectInformation Theory
dc.titleStatistical mechanics of lossy compression for non-monotonic multilayer perceptrons
dc.typetext

Files

Collections

arXiv