Storage capacity of ultrametric committee machines

Article English OPEN
Neirotti, J.P. (2014)

The problem of computing the storage capacity of a feed-forward network, with L hidden layers, N inputs, and K units in the first hidden layer, is analyzed using techniques from statistical mechanics. We found that the storage capacity strongly depends on the network architecture αc ∼ (log K)1-1/2L and that the number of units K limits the number of possible hidden layers L through the relationship 2L - 1 < 2log K.
  • References (22)
    22 references, page 1 of 3

    [1] Lee T and Munford D 2003 J. Opt. Soc. Amer. 20 1434

    [2] Lee T Munford D Romero R and Lamme V 1998 Vision. Res. 38 2429

    [3] Hinton G E 2007 Trends Cog. Scs. 11 428

    [4] Bengio 2009 Fund. and Trends in Machine Learning 2 1

    [5] Amit D J 1989 Modeling brain function (Cambridge Univ. Press, New York)

    [6] Gardner E 1987 Europhys. Lett. 4 481; 1988 J. Phys. A 21 257; Gardner E and Derrida B 1988 ibid. 21 271

    [7] Seung H S Sompolinsky H and Tishby N 1992 Phys. Rev. A 45 6056

    [8] Watkin T L H Rau A and Biehl M 1993 Rev. Mod. Phys. 65 499

    [9] Barkai E Hansel D and Kanter I 1990 Phys. Rev. Lett. 65 2312

    [10] Barkai E Hansel D and Sompolinsky H 1992 Phys. Rev. A 45 4146

  • Metrics
    0
    views in OpenAIRE
    0
    views in local repository
    5
    downloads in local repository

    The information is available from the following content providers:

    From Number Of Views Number Of Downloads
    Aston Publications Explorer - IRUS-UK 0 5
Share - Bookmark