Interpretable Machine Learning for Science

Add to your list(s) Download to your calendar using vCal

• Miles Cranmer, Princeton University
• Wednesday 14 December 2022, 15:50-16:30
• Cavendish Laboratory, Pippard Lecture Theatre & TEAMS.

If you have a question about this talk, please contact Leona Hope-Coles.

All members of the Cavendish are welcome to attend the lecture in the Pippard

Would Kepler have discovered his laws if machine learning had been around in 1609? Or would he have been satisfied with the accuracy of some black box regression model, leaving Newton without the inspiration to find the law of gravitation? In this talk I will present a review of some industry-oriented machine learning algorithms, and discuss a major issue facing their use in the natural sciences: a lack of interpretability. I will then outline several approaches I have developed with collaborators to help address these problems, based largely on a mix of structured deep learning and symbolic learning. I will walk through some applications of these techniques, and how we may gain new insights from such results.

Link to lecture:https://teams.microsoft.com/l/meetup-join/19%3ameeting_Njg0NzRjMDctYTI4Yy00MzBhLTk0MTItYWMxNmQ1YTYzNjhl%40thread.v2/0?context=%7b%22Tid%22%3a%2249a50445-bdfa-4b79-ade3-547b4f3986e9%22%2c%22Oid%22%3a%22197d44f9-a582-408b-b638-108327541246%22%7d

[Please do not select the link to the lecture until it is due to start, thank you.]

This talk is part of the Special Departmental Seminars series.

This talk is included in these lists:

• All Cavendish Laboratory Seminars
• Cavendish Laboratory, Pippard Lecture Theatre & TEAMS
• Centre for Health Leadership and Enterprise
• Featured lists
• ME Seminar
• Neurons, Fake News, DNA and your iPhone: The Mathematics of Information
• School of Physical Sciences
• Special Departmental Seminars
• Thin Film Magnetic Talks

Note that ex-directory lists are not shown.