ANC Workshop - David Sterratt

Speaker: David Sterratt

Title: Methods for generating confidence intervals on ML performance metrics

Abstract:

This is a practice for a talk giving guidance to UG4 and MSc project students on how to compare the performance of ML algorithms, and possibly for incorporation in a 2nd year undergraduate course. It's not a comprehensive overview of ML methods.

The talk is in the original spirit of ANC workshops, i.e. something of a work in progress on which I'd appreciate some feedback and comments.

If you don't know anything the topic, you may learn something; if do know something I'd appreciate your comments!