	US 12,217,190 B2
	Decision making using integrated machine learning models and knowledge graphs
Maharaj Mukherjee, Poughkeepsie, NY (US); and Utkarsh Raj, Charlotte, NC (US)
Assigned to Bank of America Corporation, Charlotte, NC (US)
Filed by Bank of America Corporation, Charlotte, NC (US)
Filed on Feb. 3, 2021, as Appl. No. 17/166,087.
Prior Publication US 2022/0245469 A1, Aug. 4, 2022
Int. Cl. G06N 20/00 (2019.01); G06N 5/022 (2023.01)

CPC G06N 5/022 (2013.01) [G06N 20/00 (2019.01)]

20 Claims

1. A computing platform comprising:

at least one processor;

a communication interface communicatively coupled to the at least one processor; and

memory storing computer-readable instructions that, when executed by the at least one processor, cause the computing platform to:

train, using historical event processing data, a machine learning model, wherein training the machine learning model comprises clustering the historical event processing data based on common characteristics, and wherein training the machine learning model configures the machine learning model to cluster input data and label the input data based on the corresponding cluster;

generate, using the historical event processing data, a knowledge graph, wherein generating the knowledge graph comprises using a depth first search method to create graph clusters configured to generate correlations between information stored in the machine learning model and additional information stored in the knowledge graph;

generate a correlation matrix, wherein the correlation matrix includes correlation values between Euclidian distances between data points of the machine learning model and hop distances between nodes of the knowledge graph;

receive new event processing data;

identify, using the machine learning model, k nearest data points corresponding to the new event processing data;

identify, using the knowledge graph, k nearest data nodes corresponding to the new event processing data;

generate, using the correlation values, first weighted relative distances between the new event processing data and the k nearest data points;

generate, using the correlation values, second weighted relative distances between the new event processing data and the k nearest data nodes;

identify, based on the first weighted relative distances and the second weighted relative distances, a data cluster for the new event processing data;

send, based on the identified data cluster, event processing information and one or more commands directing an enterprise computing device to display the event processing information, wherein sending the one or more commands directing the enterprise computing device to display the event processing information causes the enterprise computing device to display the event processing information; and

update, based on the identified data cluster and the new event processing data, the machine learning model and the knowledge graph.