Overfitting is a bad idea, let’s avoid it

I recently published a two-part post on the IBM Watson publication titled “Why Overfitting is a Bad Idea and How to Avoid It”

Part 1: Overfitting in General introduces the general topic of overfitting with a simple example and some diagrams for how to think about overfitting. Fitness concepts are introduced with two-dimensional examples for you to mentally model how higher-dimensional models work.

Part 2: Overfitting in Virtual Assistants adapts this mental model for overfitting to virtual assistants. This post helps you understand why adding more training data to your classifier is not always the best move.

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.