Self-assembling learning system and apparatus

• US 10,062,037 B2
• Filed: 08/07/2016
• Issued: 08/28/2018
• Est. Priority Date: 01/10/2006
• Status: Active Grant

• Current Assignee: Manyworlds Incorporated
• Original Assignee: Manyworlds Incorporated
• Inventors: Flinn, Steven Dennis, Moneypenny, Naomi Felina
• Primary Examiner(s): Smith, Paulinho E
• Application Number:
  US15/230,432
• Publication Number:
  US 20160342907A1
• Time in Patent Office: 751 Days
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 5/045 Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

A self-assembling learning system and apparatus determines self-assembling actions that are expected to reduce uncertainties that are embodied as probabilities and then updates the probabilities in accordance with information that results from the performing of the self-assembling actions. The updated probabilities inform the determination of subsequent self-assembling actions. Neural networks, simulations of multiple potential self-assembling actions, and expected values of information may be applied in determining the self-assembling actions that are to be performed. Sensors may be applied to receive information that inform the updating of probabilities, and the self-assembling apparatus may be a robotic device. Self-assembling actions may comprise modifications to relationships between elements of a computer-implemented system.

Claims

1. A computer-implemented method comprising:

determining automatically a first self-assembling action, wherein the first self-assembling action comprises modifying a relationship between two elements of a computer-implemented system, wherein the first self-assembling action is expected to result in a reduction of an uncertainty that is embodied by one or more probabilities;

performing automatically the first self-assembling action;

updating automatically the one or more probabilities based upon an assessment of information that results from the performing of the first self-assembling action;

determining automatically a second self-assembling action, wherein the second self-assembling action comprises modifying a relationship between two elements of the computer-implemented system, wherein the determining of the second self-assembling action is in accordance with the updated one or more probabilities; and

performing automatically the second self-assembling action.

2. The method of claim 1, further comprising:

determining automatically the first self-assembling action, wherein the first self-assembling action is determined by application of a computer-implemented neural network.

3. The method of claim 1, further comprising:

determining automatically the second self-assembling action, wherein the second self-assembling action is determined by application of a computer-implemented neural network.

4. The method of claim 1, further comprising:

updating automatically the one or more probabilities based upon the assessment of the information that results from the first self-assembling action, wherein the assessment of the information comprises performing an inference from user behavioral information associated with use of the computer-implemented system.

5. The method of claim 1, further comprising:

determining automatically the second self-assembling action in accordance with the updated one or more probabilities and an expected value of information.

6. The method of claim 5, further comprising:

determining automatically the second self-assembling action in accordance with the updated one or more probabilities, the expected value of information, and an expected cost.

7. A system comprising one or more processor-based physical devices configured to:

determine a first self-assembling action, wherein the first self-assembling action comprises modifying a relationship between two elements of the one or more processor-based devices, wherein the first self-assembling action is expected to result in a reduction of an uncertainty that is embodied by one or more probabilities;

perform the first self-assembling action;

update the one or more probabilities based upon an assessment of information that results from the performing of the first self-assembling action;

determine automatically a second self-assembling action, wherein the second self-assembling action comprises modifying a relationship between two elements of the one or more processor-based devices, wherein the second self-assembling action is determined in accordance with the updated one or more probabilities; and

perform automatically the second self-assembling action.

8. The system of claim 7, further comprising the one or more processor-based devices configured to:

determine the first self-assembling action, wherein the first self-assembling action is determined by application of a computer-implemented neural network.

9. The system of claim 8, further comprising the one or more processor-based devices configured to:

determine the second self-assembling action, wherein the second self-assembling action is determined by application of a computer-implemented neural network.

10. The system of claim 7, further comprising the one or more processor-based devices configured to:

determine the first self-assembling action, wherein the first self-assembling action is determined by simulating a plurality of self-assembling actions and an expected associated result of each of the plurality of self-assembling actions.

11. The system of claim 7, further comprising the one or more processor-based devices configured to:

update the one or more probabilities based upon the assessment of the information that results from the performing of the first self-assembling action, wherein the assessment of the result comprises performing an inference from user behavioral information associated with use of the one or more processor-based devices.

12. The system of claim 7, further comprising the one or more processor-based devices configured to:

determine the second self-assembling action in accordance with the updated one or more probabilities and an expected value of information.

13. The system of claim 12, further comprising the one or more processor-based devices configured to:

determine the second self-assembling action in accordance with the updated one or more probabilities, the expected value of information, and an expected cost.

14. An apparatus comprising a robotic device, the apparatus further comprising:

one or more sensors;

one or more processors configured to;

determine a first self-assembling action, wherein the first self-assembling action comprises modifying a relationship between two elements of the apparatus, wherein the first self-assembling action is expected to result in a reduction of an uncertainty that is embodied by one or more probabilities;

perform the first self-assembling action;

update the one or more probabilities by applying the one or more sensors to gather information that results from the performing of the first self-assembling action;

determine automatically a second self-assembling action, wherein the second self-assembling action comprises modifying a relationship between two elements of the apparatus, wherein the second self-assembling action is determined in accordance with the updated one or more probabilities; and

perform automatically the second self-assembling action.

15. The apparatus of claim 14, further comprising the one or more processors configured to:

determine the first self-assembling action, wherein the first self-assembling action is determined by application of a computer-implemented neural network.

16. The apparatus of claim 14, further comprising the one or more processors configured to:

determine the first self-assembling action, wherein first self-assembling action is determined by simulating a plurality of self-assembling actions and an expected associated result of each of the plurality of self-assembling actions.

17. The apparatus of claim 14, further comprising the one or more processors configured to:

determine the second self-assembling action, wherein the second self-assembling action is determined in accordance with the updated one or more probabilities and a value of information.