Learning method for neural network having discrete interconnection
strengths
    1.
    发明授权
    Learning method for neural network having discrete interconnection strengths 失效
    具有离散互连强度的神经网络的学习方法

    公开(公告)号:US5644681A

    公开(公告)日:1997-07-01

    申请号:US363597

    申请日:1994-12-23

    IPC分类号: G06N3/08 G06F15/18

    CPC分类号: G06N3/08

    摘要: A learning method for a neural network, in which at least a portion of the interconnection strength between neurons takes discrete values, includes the steps of updating an imaginary interconnection strength taking continuous values by using the quantity of update of the interconnection strength which has been calculated by using the discrete interconnection strength, and discretizing the imaginary interconnection strength so as to provide a novel interconnection strength.

    摘要翻译: 一种用于神经网络的学习方法,其中神经元中的至少一部分互连强度取离散值,包括以下步骤:通过使用已计算的互连强度的更新量来更新取连续值的虚拟互连强度 通过使用离散互连强度,并离散虚拟互连强度,以提供新颖的互连强度。

    Optical neurocomputer with dynamic weight matrix
    2.
    发明授权
    Optical neurocomputer with dynamic weight matrix 失效
    光学神经计算机与动态权重矩阵

    公开(公告)号:US5220642A

    公开(公告)日:1993-06-15

    申请号:US513602

    申请日:1990-04-24

    CPC分类号: G06N3/0675

    摘要: An optical computer for performing product and/or sum operations on input vector information and matrix information includes an array of light emitting elements, the array having a plurality of columns and a plurality of rows. Input circuitry is provided for supplying input vector information to the array while matrix circuitry is provided for dynamically applying matrix information to the array. Logic circuitry is provided for performing one of a logical product and a logical sum operation on the input vector information and the matrix information and for generating an output thereof.

    摘要翻译: 用于对输入矢量信息和矩阵信息执行产品和/或和操作的光学计算机包括发光元件阵列,该阵列具有多个列和多个行。 提供输入电路用于向阵列提供输入矢量信息,同时提供矩阵电路用于将矩阵信息动态地应用于阵列。 提供逻辑电路用于对输入矢量信息和矩阵信息执行逻辑积和逻辑和运算中的一个并产生其输出。