Assured AI/ML: A Statistical Methodology for Validating Training Data Quality
Open AccessAs Artificial Intelligence and Machine Learning (AI/ML) becomes more ubiquitous in everyday life – present in everything from e-mail spam filters to self-driving cars – the need to assure its accuracy, security, and validity becomes ever more critical. Traditional methods of Machine Learning assurance focus on making the models themselves more robust to uncertainty or protecting the infrastructure the models reside on from external attacks. A significant gap exists in assuring the quality of the training data that Machine Learning models use to make predictions.This praxis proposes a model-agnostic statistical method for assuring the quality of ML classification training data early in the development process before any predictions take place.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.