EMG | Dr. Simon Pepin lehalleur | Singular learning theory I: Bayesian statistics
EMG | Dr. Simon Pepin lehalleur | Singular learning theory I: Bayesian statistics Speaker: Dr Simon Pepin lehalleur (Universiteit van Amsterdam) Date: 27th Mar 2024 - 14:00 to 15:00 Venue: INI Seminar Room 2 Title: Singular learning theory I: Bayesian statistics Event: (EMG) New equivariant methods in algebraic and differential geometry Abstract: The asymptotic performance of many statistical models is controlled byproperties of the singularities of certain associated real analyticfunctions. The goal of these two lectures is to give this idea aprecise mathematical form (due to Sumio Watanabe) in the context ofBayesian statistics, and to explain recent applications to theinterpretability problem of machine learning models. In the firstlecture, I will explain Watanabe's central results, which connect themain quantities of Bayesian statistical learning theory (posteriordistribution, free energy, generalization error) to the reallog-canonical threshold of the relative entropy between the model andthe data-generating process. This is a generalization of classicalresults like the Bernstein-Von Mises theorem, which require Fisherregularity assumptions not satisfied for typical machine learningmodels. I will start with a quick overview of Bayesian statistics andof real log-canonical thresholds. 👉 Want to learn more about the talks in this stream? Seminar Link:https://www.newton.ac.uk/seminar/42502/ Programme: https://www.newton.ac.uk/event/emg/ ------------------- FOLLOW US 🌐| Website: https://www.newton.ac.uk 🎥| Main Channel: / @isaacnewtoninstitute 🐦| Twitter: / newtoninstitute 💬| Facebook: / newton.institute 📷| Instagram: / isaacnewtoninstitute 🔗| LinkedIn: / isaac-newton-institute-for-mathematical-sc... SEMINAR ROOMS 🥇| INI Seminar Room 1: / @iniseminarroom1 🥈| INI Seminar Room 2: / @iniseminarroom2 🛰️| INI Satellite Events: / @inisatellite ABOUT The Isaac Newton Institute is a national and international visitor research institute. It runs research programmes on selected themes in mathematics and the mathematical sciences with applications over a wide range of science and technology. It attracts leading mathematical scientists from the UK and overseas to interact in research over an extended period. 👉 Learn more about us and our events here: https://www.newton.ac.uk

31 - Singular Learning Theory with Daniel Murfet

Questions for Theory in the New Age of Machine Learning

Prof. Igor Wigman | Differences between Robin and Neumann eigenvalues

Bayesian Inference 1 - Zoubin Ghahramani - MLSS 2013 Tübingen

Larry Wasserman - Problems With Bayesian Causal Inference

The problem with pretending quantum mechanics makes sense | Sean Carroll

Conan O’Brien Mocks Trump At Harvard Commencement | Crowd Erupts During Viral Speech

But what is the Central Limit Theorem?
![Yann LeCun's $1B Bet Against LLMs [Part 1]](https://i.ytimg.com/vi/kYkIdXwW2AE/hqdefault.jpg?sqp=-oaymwEjCNACELwBSFryq4qpAxUIARUAAAAAGAElAADIQj0AgKJDeAE=&rs=AOn4CLDbV4izF3i-wxevCVIn7FJjoy1vlA)
Yann LeCun's $1B Bet Against LLMs [Part 1]

Free Event: Power BI Beginner to Pro 2026 Edition - Full Hands-On Tutorial

AlphaFold - The Most Useful Thing AI Has Ever Done

Billionaire's WARNING: I'm SELLING. The Crash Is Already Here!

Bayes theorem, the geometry of changing beliefs

What do tech pioneers think about the AI revolution? - The Engineers, BBC World Service

How are holograms possible?

🩸Phlebotomy Certification Practice Test – 50 Questions to Help You PASS!

Training Sand to Think: Artificial General Intelligence & Future of Physics

How AI Cracked the Protein Folding Code and Won a Nobel Prize

Group theory, abstraction, and the 196,883-dimensional monster

