US 12,462,309 B2
System and method for machine learning modeling and estimation of loss
Brian Theodore Bien, Northville, MI (US); and Alexander Douglas Findlater, Linden, MI (US)
Assigned to Insurance Quantified, LLC, Dover, DE (US)
Filed by Two Sigma Insurance Quantified, LP, New York, NY (US)
Filed on Nov. 9, 2023, as Appl. No. 18/505,759.
Claims priority of provisional application 63/424,458, filed on Nov. 10, 2022.
Prior Publication US 2024/0161204 A1, May 16, 2024
Int. Cl. G06Q 40/08 (2012.01); G06N 20/00 (2019.01)
CPC G06Q 40/08 (2013.01) [G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
at least one computer hardware processor; and
at least one non-transitory computer readable storage medium, storing processor-executable instructions, that, when executed by the at least computer hardware processor causes the at least one computer hardware processor to perform a method comprising:
obtaining one or more digital documents including insurance claims data related to a subject insurance claim;
using a first machine learning model, extracting a plurality of insurance claim features from the insurance claims data of the one or more digital documents;
analyzing the insurance claims data and insurance claims features using a plurality of machine learning models trained on a plurality of insurance loss claim records, the plurality of machine learning models comprising:
at least one machine learning model trained on open claims records to analyze a first set of insurance claims data and features associated with the subject insurance claim; and
at least one machine learning model trained on closed claims records to analyze a second set of insurance claims data and features associated with the subject insurance claim, the second set of insurance claims data being different than the first set of insurance claims data;
predicting a loss for a subject insurance claim based on outputs of the at least one machine learning model trained on open claims records and the at least one machine learning model trained on closed claims records, for later review;
generating a deliverable data product using the predicted loss, insurance claims data, and/or extracted insurance claims features; and
providing the deliverable data product to one or more external systems.