Tambe Ankit Sampat

From Drift Detection to Diagnosis: An Explainable Framework for Deployed Machine Learning

A unified drift-aware framework for explainable diagnostics in deployed machine learning systems is proposed that can support proactive model monitoring and assist practitioners in deciding whether continued monitoring, investigation, recalibration, or model retraining is required.