COM-406: Foundations of Data Science

Résumé

We discuss a set of topics that are important for the understanding of modern data science but that are typically not taught in an introductory ML course. In particular we discuss fundamental ideas and techniques that come from probability, information theory as well as signal processing.

Source officielle

https://edu.epfl.ch/coursebook/fr/foundations-of-data-science-COM-406

Page Moodle

http://go.epfl.ch/COM-406

À propos de ce résultat

Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.

Enseignants (2)

Séances de ce cours (89)

Concepts associés (976)

Detection theory or signal detection theory is a means to measure the ability to differentiate between information-bearing patterns (called stimulus in living organisms, signal in machines) and random patterns that distract from the information (called noise, consisting of background stimuli and random activity of the detection machine and of the nervous system of the operator). In the field of electronics, signal recovery is the separation of such patterns from a disguising background.

En traitement des signaux, la quantification est le procédé qui permet d'approcher un signal continu par les valeurs d'un ensemble discret d'assez petite taille. On parle aussi de quantification pour approcher un signal à valeurs dans un ensemble discret de grande taille par un ensemble plus restreint. L'application la plus courante de la quantification est la conversion analogique-numérique mais elle doit le développement de sa théorie aux problèmes de quantification pour la compression de signaux audio ou .

Dans la discipline de l'analyse numérique des mathématiques, une matrice creuse est une matrice contenant beaucoup de zéros. Conceptuellement, les matrices creuses correspondent aux systèmes qui sont peu couplés. Si on considère une ligne de balles dont chacune est reliée à ses voisines directes par des élastiques, ce système serait représenté par une matrice creuse. Au contraire, si chaque balle de la ligne est reliée à toutes les autres balles, ce système serait représenté par une matrice dense.

vignette| : un des plus vieux fragments des Éléments d'Euclide qui montre une démonstration mathématique. En mathématiques et en logique, une démonstration est un ensemble structuré d'étapes correctes de raisonnement. Dans une démonstration, chaque étape est soit un axiome (un fait acquis), soit l'application d'une règle qui permet d'affirmer qu'une proposition, la conclusion, est une conséquence logique d'une ou plusieurs autres propositions, les prémisses de la règle.

La théorie de la démonstration, aussi connue sous le nom de théorie de la preuve (de l'anglais proof theory), est une branche de la logique mathématique. Elle a été fondée par David Hilbert au début du . Hilbert a proposé cette nouvelle discipline mathématique lors de son célèbre exposé au congrès international des mathématiciens en 1900 avec pour objectif de démontrer la cohérence des mathématiques.

Cours associés (399)

MOOCs associés (98)

Concepts associés (976)

Michael Christoph Gastpar

Michael Gastpar is a (full) Professor at EPFL. From 2003 to 2011, he was a professor at the University of California at Berkeley, earning his tenure in 2008. He received his Dipl. El.-Ing. degree from ETH Zürich, Switzerland, in 1997 and his MS degree from the University of Illinois at Urbana-Champaign, IL, USA, in 1999. He defended his doctoral thesis at EPFL on Santa Claus day, 2002. He was also a (full) Professor at Delft University of Technology, The Netherlands. His research interests are in network information theory and related coding and signal processing techniques, with applications to sensor networks and neuroscience. He is a Fellow of the IEEE. He is the co-recipient of the 2013 Communications Society & Information Theory Society Joint Paper Award. He was an Information Theory Society Distinguished Lecturer (2009-2011). He won an ERC Starting Grant in 2010, an Okawa Foundation Research Grant in 2008, an NSF CAREER award in 2004, and the 2002 EPFL Best Thesis Award. He has served as an Associate Editor for Shannon Theory for the IEEE Transactions on Information Theory (2008-11), and as Technical Program Committee Co-Chair for the 2010 International Symposium on Information Theory, Austin, TX.

Rüdiger Urbanke

Rüdiger L. Urbanke obtained his Dipl. Ing. degree from the Vienna University of Technology, Austria in 1990 and the M.Sc. and PhD degrees in Electrical Engineering from Washington University in St. Louis, MO, in 1992 and 1995, respectively. He held a position at the Mathematics of Communications Department at Bell Labs from 1995 till 1999 before becoming a faculty member at the School of Computer & Communication Sciences (I&C) of EPFL. He is a member of the Information Processing Group. He is principally interested in the analysis and design of iterative coding schemes, which allow reliable transmission close to theoretical limits at low complexities. Such schemes are part of most modern communications standards, including wireless transmission, optical communication and hard disk storage. More broadly, his research focuses on the analysis of graphical models and the application of methods from statistical physics to problems in communications. From 2000-2004 he was an Associate Editor of the IEEE Transactions on Information Theory and he is currently on the board of the series "Foundations and Trends in Communications and Information Theory." In 2017 he was President of the Information Theory Society. From 2009 till 2012 he was the head of the I&C doctoral school, in 2013 he served as Dean a. i. of I&C, and since 2016 he is the Associated Dean for teaching of I&C. He is a co-author of the book "Modern Coding Theory" published by Cambridge University Press. Awards: 2021 IEEE Information Theory Society Paper Award 2016 STOC Best Paper Award 2014 La Polysphere Teaching Award 2014 IEEE Hamming Medal 2013 IEEE Information Theory Society Paper Award 2011 MASCO Best Paper Award 2011 IEEE Koji Kobayashi Award 2009 La Polysphere Teaching Award 2002 IEEE Information Theory Society Paper Award Fulbright Scholarship My students have won the following awards: M. Mondelli, 2021 IEEE Information Theory Paper Award M. Mondelli, EPFL Doctorate Award 2018 M. Mondelli, Patrick Denantes Award, 2017 M. Mondelli, IEEE IT Society Student Paper Award at ISIT, 2015 M. Mondelli, Dan David Prize Scholarship, 2015 H. Hassani, Inaugural Thomas Cover Dissertation Award, 2014 S. Kudekar, 2013 & 2021 IEEE Information Theory Paper Award A. Karbasi, Patrick Denantes Award, 2013 V. Venkatesan, Best Paper Award at MASCOTS, 2011 A. Karbasi, Best Student Paper Award at ICASSP, 2011 (with R. Parhizkar) A. Karbasi, Best Student Paper Award at ACM SIGMETRICS, 2010 (with S. Oh) S. Korada, ABB Dissertation Award, 2010 S. Korada, IEEE IT Society Student Paper Award at ISIT, 2009 (with E. Sasoglu) S. Korada, IEEE IT Society Student Paper Award at ISIT, 2008