Abstract A method of accelerating the training of an artificial neural network uses a computer configured as an artificial neural network with a network input and a network output, and having a plurality of interconnected units arranged in layers including an input layer and an output layer. Each unit has a multiplicity of unit inputs and a set of variables for operating upon a unit inputs to provide a unit output in the range positive 1 and negative 1. A plurality of examples are serially provided to the network input and the network output is observed. The computer is programmed with a back propagation algorithm for calculating changes to the sets of variables in response to feedback representing differences between the network output for each example and the desired output. The absolute magnitude of the product of an input and the corresponding output of a unit is calculated. The feedback to that unit is adjusted in response to absolute magnitude so that said feedback is larger with a larger absolute magnitude than with a smaller absolute magnitude.
Inventors
👤 Laurence F. Wood
Bibliographic Details
Application Number—
JurisdictionUS
Data Sourcegoogle_patents
Last Synchronized2026-10-25 19:29
CPC Classifications
G; G06; G06N; G06N3/00; G06N3/02; G06N3/10
Claims Analysis
Total Claims: 2
Claim 1
Independent Claim
177 words
A method of accelerating the training of an artificial neural network comprising the steps of: providing a computer configured as an artificial neural network with a network input and a network output, and having a plurality of interconnected units arranged in layers including an input layer and an output layer, each unit having a multiplicity of unit inputs and a set of variables of operating upon a unit inputs to provide a unit output in the range positive 1 and negative 1; serially providing a plurality of examples to said network input and observing the network output; programming said computer with a back propagation algorithm for calculating changes to said sets of variables in response to feedback representing differences between the network output for each example and the desired output; calculating the absolute magnitude of the product of an input and the corresponding output of a unit, and adjusting said feedback to that unit in response to said absolute magnitude so that said feedback is larger with a larger absolute magnitude than with a smaller absolute magnitude.
Claim 2
Independent Claim
177 words
A method of accelerating the training of an artificial neural network comprising the steps of: providing a computer configured as an artificial neural network with a network input and a network output, and having a plurality of interconnected units arranged in layers including an input layer and an output layer, each unit having a multiplicity of unit inputs and a set of variables of operating upon a unit inputs to provide a unit output in the range positive 1 and negative 1; serially providing a plurality of examples to said network input and observing the network output; programming said computer with a back propagation algorithm for calculating changes to said sets of variables in response to feedback representing differences between the network output for each example and the desired output; calculating the absolute magnitude of the product of an input and the corresponding output of a unit, and adjusting said feedback to that unit in response to said absolute magnitude so that said feedback is larger with a larger absolute magnitude than with a smaller absolute magnitude.