F1 Score: Don’t Use It For All Imbalanced Data

Author(s): Vlastimil Martinek

Originally published on Towards AI.

Benefits and pitfalls of the popular metric

Photo by Lightscape on Unsplash

This member-only story is on us. Upgrade to access all of Medium.

We all had to deal with imbalanced data at some point, where accuracy is not the best metric anymore, and we need something more robust. Should you use the F1 score as a metric instead? Let’s deep dive into where it came from and when it’s the right and wrong choice.

Our job is to create a model to classify if people are healthy or sick. We are given data about them, we’ve created multiple classification models, and it’s time to select the best one.

A common… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI