US 12,153,649 B1
Systems and method for data grafting to enhance model robustness
Ashutosh Verma, Bangalore (IN); Tyler Case, Phoenixville, PA (US); Paul Davis, Irving, TX (US); Matt Hord, Stanley, NC (US); Ananth Kendapadi, Charlotte, NC (US); Vinothkumar Venkataraman, Bangalore (IN); Rameshchandra Bhaskar Ketharaju, Telangana (IN); Yang Angelina Yang, Mountain View, CA (US); and Naveen Gururaja Yeri, Bangalore (IN)
Assigned to Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed by Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed on Dec. 22, 2021, as Appl. No. 17/645,735.
Int. Cl. G06F 18/214 (2023.01); G06F 18/20 (2023.01); G06F 18/21 (2023.01)
CPC G06F 18/2148 (2023.01) [G06F 18/217 (2023.01); G06F 18/285 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A method for mitigating deterioration of model performance, the method comprising:
detecting, by context analysis circuitry, occurrence of a triggering condition;
scheduling, by the context analysis circuitry and based on the occurrence of the triggering condition, retraining of a model;
in response to scheduling the retraining of the model, generating, by data grafting circuitry, a context-relevant training data set based on a target context vector, wherein the context-relevant training data set comprises a grafted set of historical data points corresponding to one or more context vectors determined to satisfy a predefined similarity threshold to the target context vector; and
retraining, by model training circuitry, the model using the context-relevant training data set to mitigate deterioration of performance of the model.