SYSTEMS AND METHODS FOR VERIFYING ACCURACY OF A MODEL BUILT USING SYNTHETIC DATA

Patent №

US 12,651,043

Granted

2026-06-09

Filed 2022

Owner

Wells Fargo Bank, N.A.

Lab

AI components

0

Assignment

None on record

Dataset

AIPD

Application

17812370

Systems, apparatuses, methods, and computer program products are disclosed for verifying accuracy of models trained using synthetic data. An example method includes selecting, by data analysis circuitry, an authentic data point set from an authentic dataset. The example method also includes generating, by modeling circuitry and using the authentic data point set, output data of a model having been trained using a synthetic dataset. The example method also includes determining, by model scoring circuitry, an accuracy score for the model based on the output data.

G06F 18/22G06F 18/217G06N 20/20G06N 20/00G06F 18/211G06N 3/08

Ownership

Wells Fargo Bank, N.A.

From the same owner

© 2026 NYSGPT2525 LLC