On-Line Learning with Restricted Training Sets: An Exactly Solvable Case

Simple item page
dc.creatorRae, H. C.
dc.creatorSollich, P.
dc.creatorCoolen, A. C. C.
dc.date1998-11-16
dc.date.accessioned2026-07-07T03:12:06Z
dc.date.available2026-07-07T03:12:06Z
dc.descriptionWe solve the dynamics of on-line Hebbian learning in large perceptrons exactly, for the regime where the size of the training set scales linearly with the number of inputs. We consider both noiseless and noisy teachers. Our calculation cannot be extended to non-Hebbian rules, but the solution provides a convenient and welcome benchmark with which to test more general and advanced theories for solving the dynamics of learning with restricted training sets.
dc.description19 pages, eps figures included, uses epsfig macro
dc.identifierhttps://arxiv.org/abs/cond-mat/9811231
dc.identifierhttp://arxiv.org/abs/cond-mat/9811231
dc.identifierJ. Phys. A: Math. Gen., 32: 3321-3339, 1999
dc.identifierdoi:10.1088/0305-4470/32/18/308
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/28780
dc.subjectDisordered Systems and Neural Networks
dc.subjectStatistical Mechanics
dc.titleOn-Line Learning with Restricted Training Sets: An Exactly Solvable Case
dc.typetext

Files

Collections

arXiv