Introduction to Probabilistic Graphical Models

Francis Bach - Nicolas Chopin
INRIA/ENS - ENSAE

Master recherche specialite "Mathématiques Appliquees",

Parcours M2 Mathématiques, Vision et Apprentissage (ENS Cachan), 1er semestre, 2018/2019

IMPORTANT: [ [ [ Please fill this form to be registered in the Moodle for the class.] ] ]

This year, the class will be taught in English. Homeworks and project may be submitted in English or in French.

This course provides a unifying introduction to probabilistic modelling through the framework of graphical models,together with their associated learning and inference algorithms.

Classes will take place on Wednesdays from 9am to 12pm at ENS Cachan, in Amphi Curie.

Moodle for the class. Now active!

Dates of classes

Below is the tentative schedule, with scribe notes from last years that might be updated for some lectures.

Date	Lecturer	Topics	Scribe notes
October 3th	Francis Bach	Introduction Maximum likelihood Linear regression Logistic regression Generative classification (Fisher discriminant)	Huu Dien Khue Le, Robin Benesse lecture1.pdf lecture1.zip Aymeric Reshef, Claire Vernade lecture2.pdf lecture2.zip
October 10th	Francis Bach	K-means EM Gaussian mixtures	Marie d'Autume, Jean-Baptiste Alayrac lecture3.pdf lecture3.zip
October 24th	Francis Bach	Graph theory Directed graphical models Undirected graphical models	Jaime Roquero, JieYing Wu lecture4.pdf lecture4.zip
October 31th	Francis Bach	Exponential families Information theory Gaussian variables Factor analysis	Thomas Belhalfaoui, Lénaic Chizat lecture5.pdf lecture5.zip Lucas Plaetevoet, Ismael Belghiti lecture6.pdf lecture6.zip
November 7th	Nicolas Chopin	Sum-product algorithm HMM	Pauline Luc, Mathieu Andreux lecture7.pdf lecture7.tex
November 14th	Nicolas Chopin	Approximate inference I Sampling and MCMC methods	Khalife Sammy, Maryan Morel lecture8.pdf lecture8.zip
November 21th	Nicolas Chopin	Approximate inference II: Variational inference	Basile Clément, Nathan de Lara lecture9.pdf lecture9.tex
December 5th	Nicolas Chopin	Bayesian methods Model selection	Gauthier Gidel, Lilian Besson lecture10.pdf lecture10.zip
December 12th		Final Exam

Homeworks

If you are not registered in the moodle, please send the homeworks (one pdf and one zip file for the code) to mvapgm@gmail.com

Homework 1, due October 24, 2018 [data] [pdf] [solution] [matlab code]

Homework 2, due November 7, 2018 [data] [pdf] [solution] [matlab code]

Homework 3, due December 5, 2018 [data] [pdf]

Exam with solution [pdf]

Internships and PhD programs

Here are a few offers that we have received related to the class. There are many others within the MVA internship forum.

Inria Lille
Berkeley
Northern Arizona
Louis Vuitton

Last updated: December 20th, 2018.