Computer architecture for identifying data clusters using correlithm objects and machine learning in a correlithm object processing system

Patent metadata

Field	Value
Publication number	US-2020175321-A1
Application number	US-201816208055-A
Country	US
Kind code	A1
Filing date	Dec 3, 2018
Priority date	Dec 3, 2018
Publication date	Jun 4, 2020
Grant date	—

Abstract

A device that includes a model training engine implemented by a processor. The model training engine is configured to obtain a set of data values associated with a feature vector. The model training engine is further configured to transform a first data value and a second data value from the set of data value into sub-string correlithm objects. The model training engine is further configured to compute a Hamming distance between the first sub-string correlithm object and the second sub-string correlithm object and to identify a boundary in response to determining that the Hamming distance exceeds a bit difference threshold value. The model training engine is further configured to determine a number of identified boundaries, to determine a number of clusters based on the number of identified boundaries, and to train the machine learning model to associate the determined number of clusters with the feature vector.

First claim

1 . A device, comprising:
a memory operable to store a machine learning model configured to map a set of feature vector inputs to a plurality of clusters; and a model training engine implemented by a processor operably coupled to the memory, configured to:
obtain a set of data values associated with a feature vector, wherein the set of data values comprises non-numerical values;
transform a first data value from the set of data value into a first sub-string correlithm object from a string correlithm object, wherein:
the string correlithm object comprises a plurality of sub-string correlithm objects;
each sub-string correlithm object is represented by an n-bit digital word; and
each sub-string correlithm object is adjacent in n-dimensional space to a preceding sub-string correlithm object and a subsequent sub-string correlithm object to form the string correlithm object;

transform a second data value from the set of data value into a second sub-string correlithm object in the string correlithm object;
compute a Hamming distance between the first sub-string correlithm object and the second sub-string correlithm object;
compare the Hamming distance to a bit difference threshold value, wherein the bit different threshold value indicates a maximum number of different bits to be considered a part of the same cluster;
identify a boundary between the first sub-string correlithm object and the second sub-string correlithm object in response to determining that the Hamming distance exceeds the bit difference threshold value;
determine a number of identified boundaries;
determine a number of clusters based on the number of identified boundaries;
train the machine learning model to associate the determined number of clusters with the feature vector.

2 . The device of claim 1 , wherein the model training engine is further configured to:
assign the first sub-string correlithm object to a first cluster; train the machine learning model with a mapping between the first sub-string correlithm object and the first cluster; assign the second sub-string correlithm object to a second cluster in response to determining that the Hamming distance exceeds the bit difference threshold value, wherein the second cluster is different than the first cluster; and train the machine learning model with a mapping between the second sub-string correlithm object and the second cluster.

3 . The device of claim 1 , wherein the model training engine is further configured to:
assign the first sub-string correlithm object to a first cluster; train the machine learning model with a mapping between the first sub-string correlithm object and the first cluster; assign the second sub-string correlithm object to the first cluster in response to determining that the Hamming distance does not exceed the bit difference threshold value; and train the machine learning model with a mapping between the second sub-string correlithm object and the first cluster.

4 . The device of claim 1 , wherein computing the Hamming distance comprises:
performing an XOR operation between the first sub-string correlithm object and the second sub-string correlithm object to generate a binary string; and counting the number of logical high values in the binary string.

5 . The device of claim 1 , wherein the model training engine is configured to normalize the set of data values.

6 . The device of claim 1 , wherein the set of data values comprises text.

7 . The device of claim 1 , wherein transforming the first data value into the first sub-string comprises:
identifying a reference sub-string correlithm object linked with a reference data value; determining a difference between the reference data value and the first data value; determining a distance parameter in the n-dimensional space based on the difference; and modifying the reference sub-string correlithm object to generate the first sub-string correlithm object, wherein modifying the reference sub-string correlithm object comprises changing a number of bits equal to the distance parameter.

8 . A machine learning model training method, comprising:
obtaining, by a model training engine implemented by a processor, a set of data values associated with a feature vector, wherein the set of data values comprises non-numerical values; transforming, by the model training engine, a first data value from the set of data value into a first sub-string correlithm object from a string correlithm object, wherein:
the string correlithm object comprises a plurality of sub-string correlithm objects;
each sub-string correlithm object is represented by an n-bit digital word; and
each sub-string correlithm object is adjacent in n-dimensional space to a preceding sub-string correlithm object and a subsequent sub-string correlithm object to form the string correlithm object;
transforming, by the model training engine, a second data value from the set of data value into a second sub-string correlithm object in the string correlithm object; computing, by the model training engine, a Hamming distance between the first sub-string correlithm object and the second sub-string correlithm object; comparing, by the model training engine, the Hamming distance to a bit difference threshold value, wherein the bit different threshold value indicates a maximum number of different bits to be considered a part of the same cluster; identifying, by the model training engine, a boundary between the first sub-string correlithm object and the second sub-string correlithm object in response to determining that the Hamming distance exceeds the bit difference threshold value; determining, by the model training engine, a number of identified boundaries; determining, by the model training engine, a number of clusters based on the number of identified boundaries; training, by the model training engine, a machine learning model to associate the determined number of clusters with the feature vector, wherein the machine learning model is configured to map a set of feature vector inputs to a plurality of clusters.

9 . The method of claim 8 , further comprising:
assigning, by the model training engine, the first sub-string correlithm object to a first cluster; training, by the model training engine, the machine learning model with a mapping between the first sub-string correlithm object and the first cluster; assigning, by the model training engine, the second sub-string correlithm object to a second cluster in response to determining that the Hamming distance exceeds the bit difference threshold value, wherein the second cluster is different than the first cluster; and training, by the model training engine, the machine learning model with a mapping between the second sub-string correlithm object and the second cluster.

10 . The method of claim 8 , further comprising:
assigning, by the model training engine, the first sub-string correlithm object to a first cluster; training, by the model training engine, the machine learning model with a mapping between the first sub-string correlithm object and the first cluster; assigning, by the model training engine, the second sub-string correlithm object to the first cluster in response to determining that the Hamming distance does not exceed the bit difference threshold value; and training, by the model training engine, the machine learning model with a mapping between the second sub-string correlithm object and the first cluster.

11 . The method of claim 8 , wherein computing the Hamming distance comprises:
performing an XOR operation between the first sub-string correlithm object and the second sub-string correlithm object to generate a binary str

Assignees

• Bank Of America

Inventors

• Panging Pankaj
• Lawrence Patrick N

Classifications

• G06F9/30029

  Logical and Boolean instructions, e.g. XOR, NOT · CPC title

• G06K9/6215

  Physics · mapped topic

• G06N20/00 Primary

  Machine learning · CPC title

• G06K9/6232

  Physics · mapped topic

• G06K9/6223 Primary

  Physics · mapped topic