Neuron unit
First Claim
1. A neuron unit comprising:
- a synapse section including;
synaptic weight generating means for generating a synaptic weight corresponding to the magnitude of an input signal, andmultiplier means for multiplying said input signal and said synaptic weight, anda neuron section including;
first adder means for adding output signals of said multiplier means together,threshold value generating means for generating a threshold value,second adder means for adding an output of said threshold value generating means to an output of said first adder means, andconversion means for subjecting an output of said second adder means to a predetermined conversion.
1 Assignment
0 Petitions
Accused Products
Abstract
A neuron unit comprises a synapse section in which input signals from external input means or preceding neuron units and synaptic weights are multiplied, and a neuron section in which output signals of the synapse section are added, subjected to thresholding, subjected to linear or nonlinear conversion, and then outputted to the outside. According to an aspect of the present invention, a synaptic weight used in the synapse section and/or a threshold used in the neuron section, which corresponds in magnitude to an external input, is obtained through interpolation from synaptic weights which have been learned previously, varying in magnitude according to external inputs. According to the other aspect of the present invention, the synaptic weight and/or the threshold is obtained in accordance with a previously set function of an external input.
26 Citations
35 Claims
-
1. A neuron unit comprising:
-
a synapse section including; synaptic weight generating means for generating a synaptic weight corresponding to the magnitude of an input signal, and multiplier means for multiplying said input signal and said synaptic weight, and a neuron section including; first adder means for adding output signals of said multiplier means together, threshold value generating means for generating a threshold value, second adder means for adding an output of said threshold value generating means to an output of said first adder means, and conversion means for subjecting an output of said second adder means to a predetermined conversion. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19)
-
-
20. A neuron unit comprising:
-
a synapse section including; storage means for storing, in advance, a plurality of synaptic weight groups, each including combinations of a plurality of synaptic weights that take discrete values resulting from learning ; interpolation means for reading out from said storage means one of said plurality of synaptic weight groups which corresponds to an input signal received from an external input means or a preceding neuron unit, and interpolating said one of the synaptic weight groups in order to obtain an interpolated value corresponding to said input signal, to thereby obtain a new synaptic weight; multiplier means for multiplying said synaptic weights obtained by said interpolation means and said input signal; and a neuron section including; first adder means for receiving output signals from said multiplier means and for adding up said output signals; threshold value generating means for generating a threshold value; second adder means for adding said threshold value to an output of said first adder means; and conversion means for subjecting an output of said second adder means to a predetermined conversion. - View Dependent Claims (21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35)
-
Specification