Analysis of Mixed Concept Drift Detectors in Deployed Machine Learning Models

Label-independent concept drift detectors represent an emerging topic in machine learning research, especially in models deployed in a production environment where obtaining labels can become increasingly difficult and costly. Concept drift refers to unforeseeable changes in the distribution of data streams, which directly impact the performance...