Towards Reliable MLOps with Drift Detectors

Simona Maggio
Jul 16, 2020 · 10 min read

Data is constantly changing, as the world from which it is collected. When a model is deployed in production, detecting changes and anomalies in new incoming data is key to make sure the predictions obtained are valid and can be safely consumed. So, how do we detect data drift?

A drifting boat in a photo by Johannes Plenio on Unsplash

In this article you will learn about two techniques used to assess whether dataset drift is occurring: the…