Computer architecture for training a correlithm object processing system

Patent metadata

Field	Value
Publication number	US-11113630-B2
Application number	US-201815927584-A
Country	US
Kind code	B2
Filing date	Mar 21, 2018
Priority date	Mar 21, 2018
Publication date	Sep 7, 2021
Grant date	Sep 7, 2021

Abstract

A correlithm object processing system that includes a trainer configured to send a node entry request to a node engine that triggers the node engine to generate an entry in a node table. The trainer is further configured to receive a source correlithm object and a target correlithm object in response to sending the node entry request. The trainer is further configured to send a real world input value and the source correlithm object to a sensor engine which triggers the sensor engine to generate an entry in a sensor table linking the real world input value and the source correlithm object. The trainer is further configured to send a real world output value and the target correlithm object to an actor engine which triggers the actor engine to generate an entry in an actor table linking the real world output value and the target correlithm object.

First claim

The invention claimed is:

1. A system configured to train a correlithm object processing system, comprising:
a node linked with a node table that identifies:
a plurality of source correlithm objects, wherein each source correlithm object is a point in an n-dimensional space represented by a binary string; and
a plurality of target correlithm objects, wherein:
each target correlithm object is a point in the n-dimensional space represented by a binary string, and
each target correlithm object is linked with a source correlithm object from among the plurality of source correlithm objects; and

a trainer operably coupled to the memory, configured to:
receive a real world input value and a real world output value;
send a node entry request to the node in response to receiving the real world input value and the real world output value;
receive a source correlithm object and a target correlithm object in response to sending the node entry request;
send the real world input value and the source correlithm object to a sensor;
send the real world output value and the target correlithm object to an actor;

the sensor operably coupled to the trainer, configured to:
receive the real world input value and the source correlithm object;
generate an entry in a sensor table linking the real world input value and the source correlithm object; and

the actor operably coupled to the trainer, configured to:
receive the real world output value and the target correlithm object;
generate an entry in an actor table linking the real world output value and the target correlithm object.

2. The system of claim 1 , wherein the node is configured to:
receive an input correlithm object linked with the real world input value;
compare the input correlithm object to the source correlithm objects in the node table; and
determine the input correlithm object does not match any of the source correlithm objects; and
wherein the trainer receives the real world input value and the real world output value after the node determines that the input correlithm object does not match any of the source correlithm objects.

3. The system of claim 1 , wherein the node is configured to:
receive an input correlithm object linked with the real world input value;
determine distances between the input correlithm object and each of the source correlithm objects in the node table, wherein the distance between the input correlithm object and a source correlithm object is based on the differences between a binary string representing the input correlithm object and binary strings linked with each of the source correlithm objects; and
determine none of the distances are within a core distance threshold; and
wherein the trainer receives the real world input value and the real world output value after the node determines that none of the distances are within the core distance threshold.

4. The system of claim 3 , wherein determining distances between the input correlithm object and each of the source correlithm objects comprises determining a hamming distance between the input correlithm object and a source correlithm object.

5. The system of claim 3 , wherein determining distances between the input correlithm object and each of the source correlithm objects comprises:
performing an XOR operation between the input correlithm object and a source correlithm object to generate a binary string; and
counting the number of logical high values in the binary string, wherein the number of logical high represents a distance.

6. The system of claim 1 , wherein the sensor is linked with a sensor table comprising:
a plurality of correlithm objects;
a plurality of real world input values; and
wherein the sensor table links each correlithm object from the plurality of correlithm objects with a real world input value from the plurality of real world input values.

7. The system of claim 1 , wherein the actor is linked with an actor table comprising:
a plurality of correlithm objects;
a plurality of real world output values; and
wherein the actor table links each correlithm object from the plurality of correlithm objects with a real world output value from the plurality of real world output values.

8. A correlithm object processing system training method, comprising:
receiving, by a trainer engine, a real world input value and a real world output value;
sending, by the trainer engine, a node entry request to a node engine in response to receiving the real world input value and the real world output value, wherein the node entry request triggers the node engine to generate an entry in a node table that identifies:
a plurality of source correlithm objects, wherein each source correlithm object is a point in an n-dimensional space represented by a binary string; and
a plurality of target correlithm objects, wherein:
each target correlithm object is a point in the n-dimensional space represented by a binary string, and
each target correlithm object is linked with a source correlithm object from among the plurality of source correlithm objects;

receiving, by the trainer engine, a source correlithm object and a target correlithm object in response to sending the node entry request;
sending, by the trainer engine, the real world input value and the source correlithm object to a sensor engine;
generating, by the sensor engine, an entry in a sensor table linking the real world input value and the source correlithm object in response to receiving the real world input value and the source correlithm object;
sending, by the trainer engine, the real world output value and the target correlithm object to an actor engine; and
generating, by the actor engine, an entry in an actor table linking the real world output value and the target correlithm object in response to receiving the real world output value and the target correlithm object.

9. The method of claim 8 , further comprising:
receiving, by a node engine, an input correlithm object linked with the real world input value;
comparing, by the node engine, the input correlithm object to the source correlithm objects in the node table; and
determining, by the node engine, the input correlithm object does not match any of the source correlithm objects; and
wherein the trainer engine receives the real world input value and the real world output value after the node engine determines that the input correlithm object does not match any of the source correlithm objects.

10. The method of claim 8 , further comprising:
receiving, by a node engine, an input correlithm object linked with the real world input value;
determining, by the node engine, distances between the input correlithm object and each of the source correlithm objects in the node table, wherein the distance between the input correlithm object and a source correlithm object is based on the differences between a binary string representing the input correlithm object and binary strings linked with each of the source correlithm objects; and
determining, by the node engine, none of the distances are within a core distance threshold; and
wherein the trainer engine receives the real world input value and the real world output value after the node engine determines that none of the distances are within the core distance threshold.

11. The method of claim 10 , wherein determining distances between the input correlithm object and each of the source correlithm objects comprises determining a hamming distance between the input correlithm object and a source correlithm object.

12. The method of claim 10 , wherein determining distances between the input correlithm object and each of the source corr

Assignees

• Bank Of America

Inventors

• Lawrence Patrick N

Classifications

• G06N7/06

  Simulation on general purpose computers · CPC title

• G06F16/902

  using more than one table in sequence, i.e. systems with three or more layers · CPC title

• G06N5/022

  Knowledge engineering; Knowledge acquisition · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• G06F9/45508

  Runtime interpretation or emulation, e g. emulator loops, bytecode interpretation · CPC title