Abstract
Systems, methods, and computer program products are provided for autonomous drift detection and model retraining. An example system includes a processor configured to aggregate, using at least one agent, event data. The processor is configured to detect, using the at least one agent and based on the event data, a set of data features exhibiting drift and, based on the set of data features, determine a set of impact scores associated with model performance. The processor is configured to predict, using the at least one agent and based on the impact scores, at least one performance outcome. The processor is configured to generate, using the at least one agent and based on the at least one performance outcome, an assessment of urgency. The processor is configured to automatically trigger, using the at least one agent and based on the assessment of urgency, retraining of the tested machine learning model.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
HITESHBHAI PATEL, KARANKUMAR S.; SAI SUMANTH YENNETI, SHANMUKHA; SUNDARJI MAJITHIYA, HARISHKUMAR; and PETER HAMILTON, SAM, "SYSTEM, METHOD, AND COMPUTER PROGRAM PRODUCT FOR AUTONOMOUS DRIFT DETECTION AND MODEL RETRAINING", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12703