Minimum Description Length

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

• Carl E. Rasmussen; Niki Kilbertus
• Thursday 25 January 2018, 13:30-15:00
• Engineering Department, CBL Seminar Room 4-38.

If you have a question about this talk, please contact Alessandro Davide Ialongo.

Abstract

“The Minimum Description Length (MDL) principle is a method for inductive inference that provides a generic solution to the model selection problem, and, more generally to the overfitting problem.” (Peter Grünwald) In this talk we will provide a very basic introduction to the philosophy and general idea behind the MDL principle that views learning as data compression. While we focus mostly on the high level goals of MDL and how it compares to, e.g., Bayesian inference or the information bottleneck, we will gently introduce a “crude two-part” version of MDL in some detail. Finally we briefly outline a “refined” version of MDL and discuss its pros and cons.

Recommended Reading

Since we start from scratch, no reading is required. For a fruitful discussion about MDL vs. Bayesian Inference vs. Frequentist at the end of the talk you can browse (the beginnings of) section 17.1 and 17.2 here:

https://homepages.cwi.nl/~pdg/book/chapter17.pdf

This talk is part of the Machine Learning Reading Group @ CUED series.

This talk is included in these lists:

• All Talks (aka the CURE list)
• Cambridge Centre for Data-Driven Discovery (C2D3)
• Cambridge Forum of Science and Humanities
• Cambridge Language Sciences
• Cambridge University Engineering Department Talks
• Cambridge talks
• Centre for Smart Infrastructure & Construction
• Chris Davis' list
• Computational Continuum Mechanics Group Seminars
• Engineering Department, CBL Seminar Room 4-38
• Featured lists
• Guy Emerson's list
• Hanchen DaDaDash
• Inference Group Journal Clubs
• Inference Group Summary
• Information Engineering Division seminar list
• Interested Talks
• ML
• Machine Learning Reading Group
• Machine Learning Reading Group @ CUED
• Machine Learning Summary
• Quantum Matter Journal Club
• Required lists for MLG
• School of Technology
• Simon Baker's List
• TQS Journal Clubs
• Trust & Technology Initiative - interesting events
• bld31
• custom
• ndk22's list
• ob366-ai4er
• rp587
• yk373's list

Note that ex-directory lists are not shown.