Device system for the autonomous generation of useful information

• US 6,356,884 B1
• Filed: 07/02/1999
• Issued: 03/12/2002
• Est. Priority Date: 10/13/1994
• Status: Expired due to Term

• Current Assignee: Stephen L. Thaler
• Original Assignee: Stephen L. Thaler
• Inventors: Thaler, Stephen L.
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): STARKS, WILBERT L
• Application Number:
  US09/347,210
• Time in Patent Office: 984 Days
• Field of Search: 706/15, 706/16, 706/28, 706/30
• US Class Current: 706/16
• CPC Class Codes: G06E 1/00 Devices for processing exclusively digital data
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology
• CPC Class Codes: G06N 3/042 Knowledge-based neural networks; Logical representations of neural networks

Abstract

An artificial neural network-based system and method for determining desired concepts and relationships within a predefined field of endeavor, including a neural network portion, which neural network portion includes an artificial neural network that has been previously trained in accordance with a set of given training exemplars, a monitor portion associated with the neural network portion to observe the data outputs produced by the previously trained artificial neural network, and a perturbation portion for perturbing the neural network portion to effect changes, subject to design constraints of the artificial neural network that remain unperturbed, in the outputs produced by the neural network portion, the perturbation portion operable such that production of an output by the neural network portion thereafter effects a perturbation of the neural network portion by the perturbation portion, the monitor portion responsive to detection of the data outputs being produced by the previously trained neural network, whereby the system is operable to derive over a period of time a plurality of input/perturbation/output mapping relationships that differ from the input/perturbation/mapping relationships of the training exemplars.

Claims

1. An artificial neural network-based system for determining, for a specified knowledge domain in a given field of endeavor as represented in a neural network, desired concepts and relationships within such predefined field of endeavor, comprising

a neural network portion having an output portion at which data outputs are produced, said neural network portion including an artificial neural network that has an input portion and which is operable to effect production of a data output from said output portion of said neural network portion when an input pattern is supplied to said artificial neural network at the input portion thereof, said artificial neural network having been previously trained in accordance with training exemplars in a given predefined field of endeavor to establish a particular knowledge domain therein and being normally operable in accordance with the constraints embodied in its design and the established knowledge domain to produce standard data outputs in response to input patterns supplied to said previously trained artificial neural network at the input portion thereof,

a monitor portion associated with said neural network portion to observe data outputs produced at the output portion of neural network portion, and

a network perturbation portion for perturbing said neural network portion to effect changes, subject to constraints embodied in the design of the previously trained artificial neural network that remain unperturbed, in the data outputs produced by said neural network portion at the output portion of said neural network portion,

said network perturbation portion operable such that production of a data output by said neural network portion thereafter effects a perturbation by said network perturbation portion of said neural network portion, such perturbation driving an operation of said artificial neural network to effect production of a data output from said neural network portion, the data output so produced establishing, based in part upon the particular varied perturbation effected, an input-perturbation-output mapping relationship within said predefined field of endeavor,

said monitor portion operable to detect and to identify, from among the data outputs being produced over a period of time at the output portion of said neural network portion when said neural network portion is so perturbed, data outputs which satisfy certain predefined criteria as preselected by a user, identification of a data output that satisfies the predefined criteria determining a desired concept within the predefined field of endeavor, which desired concept is associated with a particular input-perturbation-output mapping relationship established during operation of said system.

2. The system of claim 1 wherein the monitor portion includes a comparator portion that operates to identify from among the observed data outputs being produced at the output portion of the neural network portion certain data patterns in said observed data outputs which satisfy said predefined criteria.

3. The system of claim 2 wherein the network perturbation portion includes means for introducing internal perturbations to said previously trained artificial neural network to thereby effect a change in the operation thereof, subject to the constraints embodied in the design that remain in effect, to thereby effect production at the output portion of said neural network portion, for given input patterns being supplied to said previously trained artificial neural network, of data outputs that are distinct from the corresponding standard data outputs that would be produced in response to such given pattern of inputs by said previously trained artificial neural network in the absence of such network perturbation, which distinct data outputs remain subject to the unchanged constraints embodied in the design of said previously trained artificial neural network but identify distinct concepts in the predefined field of endeavor.

4. The system of claim 3 wherein the predefined criteria have been selected such that a data output from said neural network portion which satisfies such predefined criteria is a desired concept within the given field of endeavor.

5. The system of claim 3 wherein the monitor portion includes a program routine.

6. The system of claim 3 wherein the monitor portion comprises a second previously trained neural network.

7. The system of claim 3 wherein a subsequent perturbation of said neural network portion is determined at least in part by the response of said monitor portion to a prior data output of said neural network portion.

8. The system of claim 2 wherein the network perturbation portion includes means for introducing external perturbations to said previously trained artificial neural network to vary the particular pattern of inputs as presented at the input portion of said previously trained artificial neural network to thereby establish, upon the production of a data output in response to the particular pattern of inputs as presented, an input-output data pairing relationship.

9. The system of claim 8 wherein the predefined criteria have been selected to represent a particular desired pattern within a data output from said neural network portion, whereby, upon satisfaction of said predefined criteria, the determined input-output pairing relationship identifies a data input that results in the data output that satisfies such predefined criteria.

10. The system of claim 8 further characterized in that the monitor portion includes a program routine.

11. The system of claim 8 wherein the monitor portion comprises a second previously trained neural network.

12. The system of claim 8 wherein said means for introducing external perturbations includes a second previously trained artificial neural network and associated means for generating varied input patterns to said previously trained artificial neural network.

13. The system of claim 1 wherein said network perturbation portion is operable to effect perturbations of said neural network portion in a substantially random manner.

14. The system of claim 1 wherein said artificial neural network includes an input layer, an output layer, and at least one hidden layer.

15. The system of claim 1 including a storage medium associated with said monitor portion for retaining and storing data representative of said established input-perturbation-output mapping relationships.

16. The system of claim 1 including a system output portion associated with said monitor portion for receiving data representative of said established input-perturbation-output mapping relationships and formatting said data for use in external systems.

17. A method for determining, for a specified knowledge domain in a given field of endeavor as represented in a neural network, desired concepts and relationships within such predefined field of endeavor, comprising the steps of:

(a) providing a neural network object having an output portion at which data outputs are produced and including an artificial neural network having an input portion and which is operable to effect production of data outputs at said output portion of said neural network object in response to input patterns supplied to said input portion of said artificial neural network, the data outputs so produced having particular data patterns associated therewith, said artificial neural network having been previously trained in accordance with a set of given training exemplars in a predefined field of endeavor and having a particular established knowledge domain for the predefined field of endeavor, which previously trained neural network is normally operable in accordance with the constraints embodied in its design and its established knowledge domain to produce at said output portion, in response to input patterns supplied to the neural network at the input portion thereof, standard data outputs representative of known concepts and having associated input-perturbation-output mapping relationships,

(b) perturbing the neural network object to drive an operation of the neural network object and to effect a change in the data outputs produced thereby, such changed data outputs having input-perturbation-output mapping relationships that differ from the input-perturbation-output mapping relationships of the given training exemplars and their associated standard data outputs,

(c) monitoring the neural network object over a period of time to detect the data outputs produced thereby in response to perturbations and to identify from among the data outputs those data patterns which satisfy predefined criteria of acceptability, identification of a data pattern that satisfies the predefined criteria determining a particular concept within the predefined field of endeavor, which concept has associated therewith a particular input-perturbation-output mapping relationship.

18. The method of claim 17 wherein step (b) includes the step of:

internally perturbing said artificial neural network to thereby effect a change in the operation thereof, subject to the constraints embodied in the design that remain in effect, to thereby effect production at said output portion, for given input patterns being supplied to said previously trained artificial neural network, of data outputs that are distinct from the corresponding standard data outputs that would be produced in response to such given input patterns by said previously trained artificial neural network in the absence of such network perturbation, which distinct data outputs remain subject to the unchanged constraints embodied in the design of said previously trained artificial neural network but identify distinct concepts in the predefined field of endeavor.

19. The method of claim 17 wherein step (b) includes the step of:

introducing external perturbations to said previously trained artificial neural network to vary the particular input patterns as presented at the input portion of said previously trained artificial neural network to thereby establish, upon the production of an output in response to the particular input pattern as presented, an input-output data pairing relationship.

21. An artificial neural network-based system for obtaining, for a specified knowledge domain in a given field of endeavor as represented in a neural network, a desired concept within such predefined field of endeavor, comprising

a neural network portion including an artificial neural network object having a plurality of interrelated nodes disposed in at least an input layer, an output layer, and at least one intermediate layer, said nodes of said input layer connected to receive data inputs supplied thereto, at least one node of said neural network object connected to receive a selection criteria input, said neural network object having been previously trained in a given predefined field of endeavor, with non-restrictive selection criteria input, to establish a particular knowledge domain in said neural network object, said neural network object, after such training, operable in accordance with the constraints embodied in its design and the established knowledge domain, when a non-restrictive selection criteria input is being supplied thereto, to produce standard outputs therefrom which are representative of concepts consistent with the training of said neural network object,

a neural network perturbation portion for perturbing said artificial neural network object to drive an operation of said neural network object and to effect changes, subject to constraints embodied in the design of the artificial neural network that remain unperturbed and when a non-restrictive selection criteria input is supplied, in the outputs produced by said artificial neural network object,

said neural network object being normally operable, upon being perturbed by said neural network perturbation, when a predefined restrictive selection criteria as preselected by a user is supplied as a selection criteria input and such selection criteria is satisfied, to produce an output therefrom, which output is representative of a desired concept, and

a system output portion for receiving outputs that are representative of a desired concept in accordance with the predefined restrictive selection criteria.

20. The method of claim 21 wherein the predefined criteria have been selected to represent a particular desired pattern within a data output from said previously trained neural network, whereby, upon satisfaction of said predefined criteria, the determined input-output pairing relationship identifies a data input that results in the data output that satisfies such predefined criteria.

22. The system of claim 21 wherein said neural network object includes first and second artificial neural networks, said first artificial neural network connected to receive said data inputs and to produce data outputs in response thereto, said second artificial neural network connected to receive said predefined selection criteria and at least some of said data outputs produced by said first artificial neural network.

23. An artificial neural network-based system for determining, for a specified knowledge domain in a given field of endeavor as represented in a neural network, desired concepts and relationships within such predefined field of endeavor, comprising

a neural network portion including an artificial neural network that has an input portion and an output portion and is operable to produce outputs at said output portion when inputs are supplied to said artificial neural network at the input portion thereof, said artificial neural network having been previously trained in accordance with training exemplars in a given predefined field of endeavor to establish a particular knowledge domain therein and being normally operable in accordance with the constraints embodied in its design and the established knowledge domain to produce standard outputs in response to a pattern of inputs supplied to said previously trained artificial neural network at the input portion thereof,

a neural network perturbation portion for perturbing said neural network portion to degrade said neural network portion and the knowledge domain of said neural network to effect changes, subject to constraints embodied in the design of the artificial neural network that remain unperturbed, in the outputs produced by said neural network portion at the output portion of said neural network portion, said changed outputs differing from the standard outputs that would be produced by said neural network portion if the knowledge domain of said neural network were not degraded, each output so produced establishing an input-perturbation-output mapping relationship within said predefined field of endeavor, and

a monitor portion associated with said neural network portion to observe the outputs produced at the output portion of the artificial neural network over a period of time and to identify from among said outputs so produced particular outputs that satisfy predefined criteria as preselected by a user, identification of an output that satisfies the predefined criteria determining a desired concept within the predefined field of endeavor, which desired concept is associated with a particular input-perturbation-output mapping relationship established during operation of said system.