US5734796A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Self-organization of pattern data with dimension reduction through learning of non-linear variance-constrained mapping
Assignee: CA Inc
•
Filing: 1995-09-29
•
Granted: 1998-03-31
•
Expires: 2015-09-29
Abstract
Abstract The subject system provides a self-organized reduced-dimension remapping of pattern data. The system functions to a mapping from an original pattern space to a reduced-dimension space in an unsupervised nonlinear manner, but with a constraint that the overall variance in a representation of the data be conserved. This approach relates to but is different from both the Karhuren-Loeve (K-L) transform and auto-associative approaches which emphasize feature extraction, and also from the Advanced Reasoning Tool (ART) and feature mapping approaches which emphasize category formation based on similarity in the original representation. The subject system is highly efficient computationally. The reduced-dimension representation is suitably further simplified with ART or feature mapping techniques, as appropriate and as desired.
Inventors
- 👤 Yoh Han Pao
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-01 21:51
CPC Classifications
G; G06; G06N; G06N3/00; G06N3/02; G06N3/08; G06N3/084
Claims Analysis
Total Claims: 11
Claim 1
Independent Claim
253 words
A system for organization of multi-dimensional pattern data into a two-dimensional representation comprising: a neural network comprised of a plurality of layers of nodes, the plurality of layers including: an input layer composed of a plurality of nodes, and an output layer composed of first and second non-linear nodes sharing a common internal network representation; receiving means for receiving multi-dimensional pattern data into the input layer of the neural network; means for generating an output signal at the output layer of the neural network corresponding to received multi-dimensional pattern data; and training means for completing a training of the neural network, wherein the training means includes means for completing the training of the neural network via backpropagation for progressively changing weights for nodes of the output layer, the receiving means including means for receiving the multi-dimensional pattern data into the input layer of the neural network as a set {xp} of data pattern vectors into an input layer of a neural network, wherein p=1,2, . . . ,P and wherein P is defined as a positive integer, and wherein the set of data pattern vectors has a total variance defined as, ##EQU5## wherein <x i > denotes at least one of the average or mean of x i ; and the training means further includes, means for training the neural network by backpropagation by progressively changing weights w at an output layer of the neural network composed of first and second non-linear nodes sharing a common internal network representation in accordance with, ##EQU6##
Claim 2
Independent Claim
253 words
A system for organization of multi-dimensional pattern data into a two-dimensional representation comprising: a neural network comprised of a plurality of layers of nodes, the plurality of layers including: an input layer composed of a plurality of nodes, and an output layer composed of first and second non-linear nodes sharing a common internal network representation; receiving means for receiving multi-dimensional pattern data into the input layer of the neural network; means for generating an output signal at the output layer of the neural network corresponding to received multi-dimensional pattern data; and training means for completing a training of the neural network, wherein the training means includes means for completing the training of the neural network via backpropagation for progressively changing weights for nodes of the output layer, the receiving means including means for receiving the multi-dimensional pattern data into the input layer of the neural network as a set {xp} of data pattern vectors into an input layer of a neural network, wherein p=1,2, . . . ,P and wherein P is defined as a positive integer, and wherein the set of data pattern vectors has a total variance defined as, ##EQU5## wherein <x i > denotes at least one of the average or mean of x i ; and the training means further includes, means for training the neural network by backpropagation by progressively changing weights w at an output layer of the neural network composed of first and second non-linear nodes sharing a common internal network representation in accordance with, ##EQU6##
Claim 3
Dependent Claim
37 words
A system according to claim 1, wherein said neural network further comprises a hidden layer composed of hidden nodes, wherein weights w for each hidden node of the neural network is progressively improved in accordance with, ##EQU7##
Claim 4
Independent Claim
247 words
A system for organizing multi-dimensional pattern data into a reduced dimensional representation comprising: a neural network comprised of a plurality of layers of nodes, the plurality of layers including: an input layer comprised of a plurality of input nodes, and an output layer comprised of a plurality of non-linear output nodes sharing a common internal network representation, wherein the number of non-linear output nodes is less than the number of input nodes; receiving means for receiving multi-dimensional pattern data into the input layer of the neural network; means for generating an output signal at the output layer of the neural network corresponding to received multi-dimensional pattern data; and training means for completing a training of the neural network, wherein the training means includes means for completing the training of the neural network via backpropagation for progressively changing weights for the output nodes, the receiving means includes means for receiving the multi-dimensional pattern data into the input layer of the neural network as a set {xp} of data pattern vectors, wherein p=1,2, . . . ,P and wherein P is defined as a positive integer, and wherein the set of data pattern vectors has a total variance defined as, ##EQU8## wherein <x i > denotes at least one of the average or mean of x i ; and the training means further includes, means for training the neural network by backpropagation by progressively changing weights w at the output layer of the neural network in accordance with, ##EQU9##
Claim 5
Independent Claim
247 words
A system for organizing multi-dimensional pattern data into a reduced dimensional representation comprising: a neural network comprised of a plurality of layers of nodes, the plurality of layers including: an input layer comprised of a plurality of input nodes, and an output layer comprised of a plurality of non-linear output nodes sharing a common internal network representation, wherein the number of non-linear output nodes is less than the number of input nodes; receiving means for receiving multi-dimensional pattern data into the input layer of the neural network; means for generating an output signal at the output layer of the neural network corresponding to received multi-dimensional pattern data; and training means for completing a training of the neural network, wherein the training means includes means for completing the training of the neural network via backpropagation for progressively changing weights for the output nodes, the receiving means includes means for receiving the multi-dimensional pattern data into the input layer of the neural network as a set {xp} of data pattern vectors, wherein p=1,2, . . . ,P and wherein P is defined as a positive integer, and wherein the set of data pattern vectors has a total variance defined as, ##EQU8## wherein <x i > denotes at least one of the average or mean of x i ; and the training means further includes, means for training the neural network by backpropagation by progressively changing weights w at the output layer of the neural network in accordance with, ##EQU9##
Claim 6
Dependent Claim
33 words
A system according to claim 3, wherein said neural network further comprises a hidden layer composed of hidden nodes, wherein weights w for each hidden node is progressively improved in accordance with, ##EQU10##
Claim 7
Independent Claim
165 words
A method for effecting the organization of multi-dimensional pattern data into a reduced dimensional representation using a neural network having an input layer composed of a plurality of input nodes, and an output layer composed of a plurality of non-linear output nodes sharing a common internal network representation, the number of non-linear output nodes being less than the number of input nodes, said method comprising: receiving a set {xp} of data pattern vectors into the input layer of the neural network, wherein p=1,2, . . . ,P and wherein P is defined as a positive integer, and wherein the set of data pattern vectors has a total variance defined as, ##EQU11## wherein <x i > denotes at least one of the average or mean of x i ; and training the neural network by backpropagation by progressively changing weights w at the output layer of the neural network in accordance with, ##EQU12## displaying a multi-dimensional output signal from the output layer of the neural network.
Claim 8
Independent Claim
165 words
A method for effecting the organization of multi-dimensional pattern data into a reduced dimensional representation using a neural network having an input layer composed of a plurality of input nodes, and an output layer composed of a plurality of non-linear output nodes sharing a common internal network representation, the number of non-linear output nodes being less than the number of input nodes, said method comprising: receiving a set {xp} of data pattern vectors into the input layer of the neural network, wherein p=1,2, . . . ,P and wherein P is defined as a positive integer, and wherein the set of data pattern vectors has a total variance defined as, ##EQU11## wherein <x i > denotes at least one of the average or mean of x i ; and training the neural network by backpropagation by progressively changing weights w at the output layer of the neural network in accordance with, ##EQU12## displaying a multi-dimensional output signal from the output layer of the neural network.
Claim 9
Dependent Claim
37 words
A method according to claim 5, wherein said neural network further comprises a hidden layer composed of hidden nodes, wherein weights w for each hidden node of the neural network is progressively improved in accordance with, ##EQU13##
Claim 10
Dependent Claim
16 words
A method according to claim 5, wherein said multi-dimensional output signal is a two-dimensional output signal.
Claim 11
Dependent Claim
20 words
A method according to claim 7, wherein said two-dimensional output signal includes data points plotting in relation to 2-dimensional axes.
References Cited (Backward Citations)
US5003490A
Cited
US5113483A
Cited
US5218529A
Cited
US5546503A
Cited
US5200816A
Cited
US5293456A
Cited
US5335291A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2001-03-30 | 2001-09-29 | 2002-03-31 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2005-03-30 | 2005-09-29 | 2006-03-31 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2009-03-30 | 2009-09-29 | 2010-03-31 | $7700.00 | Window Closed / Lapsed |