Sites Inria

Version française

RDV du Plateau Inria

Graphs in Machine Learning

CC0 / Pexels

On Thursday, November 9, from 9am to 10:30am at the Plateau Inria (EuraTechnologies) will take place a conference about graphs in machine learning. This subject is the expertise of project-team Sequel from our research center. Register.

  • Date : 9/11/2017
  • Place : Plateau Inria, EuraTechnologies - 165 avenue de Bretagne, Lille

Where is Justin Bieber? Graph bandits for online influence maximization

How can a message or idea be disseminated quickly - and as cheaply as possible - on social networks? How can their members be persuaded to vote for a candidate or buy a product?  

The solution is simple: target high-profile users who will take care of disseminating the information for you! But how can they be located in a social network like Facebook?  In this presentation, we will show you a method that makes it possible to find members as influential as Justin Bieber or Lady Gaga!

Meet Sequel project-team


Michal Valko - © Inria / Photo A. Wrona

Michal Valko is a junior scientist in Sequel project-team* at Inria Lille - Nord Europe, lead by Philippe Preux and Rémi Munos. He also teaches the course Graphs in Machine Learning at l'ENS Cachan. Michal is primarily interested in designing algorithms that would require as little human supervision as possible. This means 1) reducing the “intelligence” that humans need to input into the system and 2) minimising the data that humans need spend inspecting, classifying, or “tuning” the algorithms. Another important feature of machine learning algorithms should be the ability to adapt to changing environments. That is why he is working in domains that are able to deal with minimal feedback, such as bandit algorithms, semi-supervised learning, and anomaly detection.

Most recently he has worked on sequential algorithms with structured decisions where exploiting the structure can lead to provably faster learning. In the past the common thread of Michal's work has been adaptive graph-based learning and its application to the real world applications such as recommender systems, medical error detection, and face recognition. His industrial collaborators include Adobe, Intel, Technicolor, and Microsoft Research. He received his PhD in 2011 from University of Pittsburgh under the supervision of Miloš Hauskrecht and after was a postdoc of Rémi Munos.

*Sequel project-team is associated with CNRS, University of Lille − sciences et technologies et University of Lille − sciences humaines et sociales. Within the UMR 9189 CNRS-Centrale Lille-University of Lille − sciences et technologies, CRIStAL.

Keywords: Graphs Equipe-projet Sequel R&DV du Plateau Machine learning IA