Abstract The invention provides for a data processing system with a plurality of neurons each of which outputs a predetermined data after igniting according to the predetermined processing results performed onto an input data; neurons in the above neural layers are divided into the neuron groups with a fixed number of neurons; neurons in a part of each neuron group belong to the adjacent neuron group at the same time; neurons in each neuron group correspond to neurons in the following neural layer, respectively; number of neurons Nn in the nth neural layer can be calculated by the following formula: Nn = ( Nn-1 - v)/( a - v)<2> where, "Nn-1" is the number of neurons at (n-1)th layer; "a" is the number of neurons in one neuron group; and "v" is the number of strings that belong to another adjacent neuron group. <IMAGE>
Data processing system characterized in that; a plurality of neural layers is prepared which comprises a plurality of neurons each of which outputs a predetermined data after igniting according to the predetermined processing results performed onto an input data; neurons in the above neural layers are divided into the neuron groups with a fixed number of neurons; neurons in a part of each neuron group belong to the adjacent neuron group at the same time; neurons in each neuron group correspond to neurons in the following neural layer, respectively; number of neurons N n in the nth neural layer can be calculated by the following formula: N n = {(√ N n-1 ¯ - v)/(√a - v)}² where, "N n-1 " is the number of neurons at (n-1)th layer; "a" is the number of neurons in one neuron group; and "v" is the number of strings that belong to another adjacent neuron group.
Claim 2
Dependent Claim
155 words
Data processing system characterized in that: a plurality of neural layers are prepared which comprises a plurality of neurons each of which outputs a predetermined data after igniting according to the predetermined processing results performed onto an input data; a plurality of the mutually independent data processing portions are prepared which comprise a plurality of the above neural layers; neurons in a part of each neuron group belong to the adjacent neuron group at the same time; neurons in each neuron group correspond to neurons in the following neural layer, respectively; number of neurons N n in the nth neural layer can be calculated by the following formula: N n = {(√ N n-1 ¯ - v)/(√a - v)}² where, "N n-1 " is the number of neurons at (n-1)th layer; "a" is the number of neurons in one neuron group; and "v" is the number of strings that belong to another adjacent neuron group.
Claim 3
Dependent Claim
50 words
Data processing system according to claim 2, characterized in that: said system includes a preprocessing portion comprising a plurality of the mutually independent data processing portions and a pro-processing portion comprising at least one data processing portion; output data from the above preprocessing portion is input to the pro-processing portion.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.