Statistical Learning Theory, Spring Semester 2021

Instructors

Prof. Dr. Joachim M. Buhmann
Dr. Carlos Cotrini

Assistants

Dr. Paolo Penna
Ami Beuret
Evgenii Bykovetc
Joao Carvalho
Alina Dubatovka
Mikhail Karasikov
Ivan Ovinnikov

General Information

The ETHZ Course Catalogue information can be found here.

The course covers advanced methods of statistical learning. The fundamentals of Machine Learning as presented in the course "Introduction to Machine Learning" and "Advanced Machine Learning" are expanded and, in particular, the following topics are discussed:

  • Variational methods and optimization. We consider optimization approaches for problems where the optimizer is a probability distribution. We will discuss concepts like maximum entropy, information bottleneck, and deterministic annealing.
  • Clustering. This is the problem of sorting data into groups without using training samples. We discuss alternative notions of "similarity" between data points and adequate optimization procedures.
  • Model selection and validation. This refers to the question of how complex the chosen model should be. In particular, we present an information theoretic approach for model validation.
  • Statistical physics models. We discuss approaches for approximately optimizing large systems, which originate in statistical physics (free energy minimization applied to spin glasses and other models). We also study sampling methods based on these models.

Time and Place

Type Time Place
Lectures Mon 14-16, Tue 17-18 Zoom link
Exercises Mon 16-18 Zoom link

Lectures and tutorials

All lectures and tutorials take place via Zoom. To access the Zoom link, use your NETHZ credentials. A recording will be made available in the webpage within 24h after the lecture or the tutorial.

Piazza website

This term we will be using Piazza for class discussion. The system is highly catered to getting you help fast and efficiently from classmates and the teaching team. Rather than emailing questions to the teaching team, I encourage you to post your questions on Piazza. If you have any problems or feedback for the developers, email team@piazza.com.
Find our class page here. Use your NETHZ credentials to access this link.

Material

<
Date Lecture Tutorial Exercises Reference
Feb 22

Intro
Lecture 1
Video 1
Video 2
No tutorial Exercise 1
Solution 1
Mar 1

Max. entropy
Lecture 2

Exercise 2
[Scr20] Ch 2.1-2.2
Mar 8

Max. entropy

Sampling


[Scr21] Ch 1, 2
Mar 15

Deterministic annealing


[Scr21] Ch 3
Mar 22

Laplace method

Histogram clustering





[Scr21] Ch 4, 5
[Scr20] Sec 2.7
Mar 29

Param. distr. clustering

Info. bottleneck






PDC paper
IBM paper
[Scr20] Sec 2.7-2.8
Apr 12

Pairwise clustering


CSE paper
PC-DA paper
[Scr20] Ch 3
Apr 20

Mean field
No lecture on Monday
No tutorial No exercise
Apr 26
May 3

Model selection
May 10

Posterior agreement
[Scr21] Ch 10
May 17
May 24 No lecture on Monday No tutorial No exercise
May 31

Conclusion

Past written Exams:


Exam 2019
Draft Solution 2019
Exam 2020 (with solution)

Projects

Projects are coding exercises that concern the implementation of an algorithm taught in the lecture/exercise class.

There will be seven coding exercises, with a time span of approximately two weeks per coding exercise. Each one of them will be graded as not passed or with a passing grade ranging from 4 to 6. The project part is passed if the student receives a passing grade in at least four coding exercises, and in that case the grade of the project part is the average of the four best coding exercises.

In order to be admitted to the exam the student has to pass the project part, and the final grade is 0.7 exam + 0.3 project, rounded to the nearest quarter of unit. More details and info are contained into the project repository (including the dates of the various projects and instructions on how to submit solutions).

Reading

  • [Scr21] Course script. To be completed in the next 2 years.
  • [Scr20] Previous script. It's no longer maintained, but it contains useful notes for some chapters not covered yet in [Scr21].
  • Duda, Hart, Stork: Pattern Classification, Wiley Interscience, 2000.
  • Hastie, Tibshirani, Friedman: The Elements of Statistical Learning, Springer, 2001.
  • L. Devroye, L. Gyorfi, and G. Lugosi: A probabilistic theory of pattern recognition. Springer, New York, 1996

Web Acknowledgements

The web-page code is based (with modifications) on the one of the course on Machine Learning (Fall Semester 2013; Prof. A. Krause).