From the course: Machine Learning and AI Foundations: Producing Explainable AI (XAI) and Interpretable Machine Learning Solutions

Unlock the full course today

Join today to access over 24,900 courses taught by industry experts.

XAI for debugging models

XAI for debugging models - KNIME Tutorial

From the course: Machine Learning and AI Foundations: Producing Explainable AI (XAI) and Interpretable Machine Learning Solutions

Start my 1-month free trial Buy for my team

XAI for debugging models

“

- Imagine that you build your model, do cross validation and everything else you've been taught, and yet when you start to put the model into user testing the performance is awful. In a recent interview, Marco Ribeiro tells a compelling story. We'll be hearing more about that name because he's the lead author of the paper that introduced LIME, one of the techniques we'll be discussing. He explains that he didn't set out to study XAI at first. He became interested in it because he ran into a problem on a project. The model had learned to distinguish between how we had collected positive and negative data rather than picking up on what it should have been picking up on. The researcher and academic Cynthia Rudin, a name we hear repeatedly in this course, often refers to a similar case study. A medical AI model was discovered to have been detecting the presence of a word in the image rather than diagnostic clue in the image…

Contents