DEV Community

Adventures in Machine Learning

Rethinking ML Monitoring - ML 060

Mo’ advancements mean mo’ problems, and today, Michael and Ben are diving into the biggest issues of ML monitoring in 2022. They lay out the ELEVEN (cause nothing good comes easy) potential pitfalls that you should know this year, the important questions that you NEED to ask yourself before launching your baby, and this ONE phenomenon that reveals a fundamental flaw in models versus the real deal.

In This Episode

  1. The checklist of ELEVEN problems with ML monitoring that you need to consider in 2022
  2. Why THIS is inevitable with ML monitoring no matter what (and why it’s not the end of the world…unless you make it)
  3. Before you make an assumptions about your model, ask yourself THESE questions and tread carefully
  4. This ONE phenomenon that reveals a HUGE flaw in models vs. real world

Sponsors

Sponsored By:

Episode source