Control and Monitoring of Artificial Intelligence Algorithms

TL;DR

This paper emphasizes the importance of monitoring AI models after deployment, focusing on data and concept drift, and introduces metrics to evaluate model performance amid distributional changes.

Contribution

It provides a comprehensive overview of data and concept drift, and proposes metrics for monitoring AI models in dynamic environments.

Findings

Identifies key types of drift affecting AI models.

Introduces metrics for detecting performance fluctuations.

Highlights the importance of ongoing model oversight.

Abstract

This paper elucidates the importance of governing an artificial intelligence model post-deployment and overseeing potential fluctuations in the distribution of present data in contrast to the training data. The concepts of data drift and concept drift are explicated, along with their respective foundational distributions. Furthermore, a range of metrics is introduced, which can be utilized to scrutinize the model's performance concerning potential temporal variations.