Fiches de cours 2017-2018

PDF
 

Attention, ce contenu correspond à la fiche de cours de l'année passée


A network tour of data science

EE-558

Enseignant(s) :

Frossard Pascal
Vandergheynst Pierre

Langue:

English

Summary

This course offers an introduction to algorithms in data science and network analysis. A major goal is to design and analyze graph-based algorithms in the context of learning, recommendation, visualization, and representation. The course provides coding exercices on real-world cases.

Content

Context

In the last decade, our information society has mutated into a data society, where the volume of worldwide data doubles every 1.5 years. How to make sense of such tremendous volume of data? Developing effective techniques to extract meaningful information from large-scale and high-dimensional dataset has become essential for the success of business, government and science.

Objective

The goal of this course is to provide a broad introduction to effective algorithms in data science and network analysis. A major effort will be given to show that existing data analysis techniques can be defined and enhanced on graphs. Graphs encode complex structures like cerebral connection, stock exchange, and social network. Strong mathematical tools have been developed based on linear and non-linear graph spectral harmonic analysis to advance the standard data analysis algorithms. Main topics of the course are networks, unsupervised and supervised learning, recommendation, visualization, sparse representation, multi-resolution analysis, neuron network, and large-scale computing.

Structure

The course is organized into two parts: lectures (2 hours) and coding exercises (1 hour). The essential objective of the exercises is to apply the theory on real-world cases.
Evaluation
Evaluation will be conducted on a continuous basis: homeworks and coding assignments.

Bio Prof. Pierre Vandergheynst: Full professor of EE and CS. Developer of graph wavelets, a multi-resolution data analysis technique based on spectral harmonic analysis.
Bio Dr. Xavier Bresson: Scientist Researcher in EE. Developer of total variation clustering, an exact relaxation technique for graph balanced cut problems. Publications at NIPS, ICML, JMLR. Invited Research Fellow at the 2014 workshop "Network Science and Graph algorithms", ICERM, Brown, US and the 2017 Workshop on 'Variational Methods, New Optimization Techniques and Fast Numerical Algorithms', Isaac Newton Institute, Cambridge, UK.
Outline of the 14 weeks
See Annex or this link: https://www.dropbox.com/s/tpw9xd7my7374ym/outline_course_GDS.txt?dl=0 

 

Keywords

data sci mining ence, machine learning, data

Learning Prerequisites

Required courses

linear algebra, calculus, digital signal processing or equivallent

Learning Outcomes

By the end of the course, the student must be able to:

Dans les plans d'études

Semaine de référence

 LuMaMeJeVe
8-9     
9-10     
10-11BC03   DIA005
11-12   
12-13     
13-14     
14-15     
15-16     
16-17     
17-18     
18-19     
19-20     
20-21     
21-22     
 
      Cours
      Exercice, TP
      Projet, autre

légende

  • Semestre d'automne
  • Session d'hiver
  • Semestre de printemps
  • Session d'été
  • Cours en français
  • Cours en anglais
  • Cours en allemand