Seminar: Data Structures of the Future: Concurrent, Optimistic, and Relaxed Speaker by Dan Alistarh, Microsoft Research

Date: Friday December 5th 2014 at 2:00PM

Room: IXXI

Abstract: The need to process larger and larger amounts of data as efficiently as possible has been one of the main computing trends of the past decade.
This new set of requirements significantly changes the way data structures are designed, implemented and employed. In this talk, I will give a couple of examples of such new data structure designs from our work, and describe some of the challenges in the area.

Based on joint work with Nir Shavit (MIT), Keren Censor-Hillel (Technion), and Milan Vojnovic (Microsoft Research)

Short Bio:
Dan’s research focuses on building scalable algorithms, and combines theoretical ideas and practical implementations. Before Microsoft Research, Dan was an PostDoc at MIT, working with Nir Shavit and Nancy Lynch. He obtained his PhD from EPFL, under the guidance of Rachid Guerraoui. For details, please see

Seminar: Matrix Completion on Graphs by Prof. Pierre Vandergheynst (EPFL)

Date: 28/11/2014 at 2PM
Room: lecture room of the CBP (in the LR6 building of Monod campus) about :

Title: Matrix Completion on Graphs

Abstract: The problem of finding the missing values of a matrix given a few of its entries, called matrix completion, has gathered a lot of attention in the recent years. Although the problem is NP-hard, Candès and Recht showed that it can be exactly relaxed if the matrix is low-rank and the number of observed entries is sufficiently large. In this work, we introduce a novel matrix completion model that makes use of proximity information about rows and columns by assuming they form communities. This assumption makes sense in several real-world problems like in recommender systems, where there are communities of people sharing preferences, while products form clusters that receive similar ratings. Our main goal is thus to find a low-rank solution that is structured by the proximities of rows and columns encoded by graphs. We borrow ideas from manifold learning and signal processing to constrain our solution to be smooth on these graphs, in order to implicitly force row and column proximities. Our matrix recovery model is formulated as a convex non-smooth optimization problem, for which a well-posed iterative scheme is provided. We study and evaluate the proposed matrix completion on synthetic and real data, showing that the proposed structured low-rank recovery model outperforms the standard matrix completion model in many situations.

This is joint work with Vassilis Kalofolias, Michael Bronstein and Xavier Bresson.

Publication in Journal of the Royal Society Interface

“Complex contagion process in spreading of online innovation” by Márton Karsai, Gerardo Iñiguez, Kimmo Kaski and János Kertész

Published 22 October 2014 doi: 10.1098/​rsif.2014.0694 J. R. Soc. Interface 6 December 2014 vol. 11 no. 101 20140694

More information on the web of J. R. Soc. Interface

Workshop Internet Of Things / Equipex FIT IoT-LAB

Call for participation
Workshop Internet Of Things (IoT)
FIT IoT-LAB inauguration and tutorials

https://www.iot-lab.info
Montbonnot, 6 & 7 November 2014

Proposal submission
Send your proposal to conf-iotlab@inria.fr including name and affiliation of the speaker, the title of your presentation and a short description proposal.
Submission deadline : October 19, 2014

Registration
Registration on Workshop (day1) and Tutorial (day2) is free an must be sent (First Name/Last Name/email/Affiliation) to conf-iotlab@inria.fr
Registration deadline : October 24, 2014

Scope
The project -Future Internet of Things (FIT) – is one of 52 winning projects from the first wave of the French Ministry of Higher Education and Research’s “Équipements d’Excellence” (Equipex) research grant programme. In this context, a Workshop is organized on the Internet of Things and experimental platforms IoT.
This workshop will be held at the research center INRIA Grenoble Rhone-Alpes on 6 & 7 November 2014; it is supported by the FIT consortium and the chair INRIA Schneider.
This workshop has the dual purpose by allowing exchanges on research around the Internet of Things and presenting talks and tutorials on various topics on the IoT domain. It specifically target experimentation through testing, verification, deployment, integration, management and federation of experimental platforms.
Send your proposal to conf-iotlab@inria.fr including name and affiliation of the speaker, the title of your talk and a short description proposal.
The workshop will bring together people from academia & industry, in several areas of research so that they can exchange ideas about the current and the future of the IoT.
The second day will be dedicated to a tutorial on using FIT IoT-LAB.

Autumn 2014 CODDDE meeting

All informations on http://jlguillaume.free.fr/coddde/meetings.php#autumn2014

DANTE is present at ECCS 2014

Laura, Laetitia, Matteo, Marton are presenting severals papers at ECCS 2014.

Ha Duong PHAN is invited professor of ENS Lyon and UCBL during march-april 2014

Ha Duong PHAN (http://vie.math.ac.vn/~phduong/) from the Institute of Mathematics of the Vietnam Academy of Science and Technology will stay in the DANTE team of the LIP, at IXXI, as an invited professor of ENS Lyon and UCBL during march and april 2014. She will give a seminar on “Chip Firing Game: Lattice structure and recognition algorithm” on Tuesday 18th March at 15.00 in ENS Lyon, site Jacques Monod, room 115.

Ha Duong Phan, Institute of Mathematics, Vietnam Academy of Science and Technology
Tuesday 18th March at 15.00 in ENS Lyon, site Jacques Monod, room 115.

In this talk, I will present the lattice structure of the configuration space of Chip Firing Game – a discrete dynamical model introduced by Dhar (1990) and A. Björner, L. Lovász and W. Shor (1991). The class of lattices generated by Chip Firing Games (CFGs) contains strictly the class of distributive lattice and is strictly included in the class of upper locally distributive lattices (ULD). However a necessary and sufficient criterion for this class is still an open question. We will give such a criterion. This criterion provides a polynomial-time algorithm for constructing a CFG which generates a given lattice if such a CFG exists.

Paulo Gonçalves keynote speaker at CLOSER 2014

More information at CLOSER 2014 (4th International Conference on Cloud Computing and Services Science) web site: http://closer.scitevents.org/Home.aspx

Benjamin’s paper accepted at ICASSP: Semi-Supervised Learning for Graph to Signal Mapping: a Graph Signal Wiener Filter Interpretation

http://hal.inria.fr/hal-00942695/en

Abstract: In this contribution, we investigate a graph to signal mapping with the objective of analysing intricate structural properties of graphs with tools borrowed from signal processing. We successfully use a graph-based semi-supervised learning approach to map nodes of a graph to signal amplitudes such that the resulting time series is smooth and the procedure efficient and scalable. Theoretical analysis of this method reveals that it essentially amounts to a linear graph-shift-invariant filter with the a priori knowledge put into the training set as input. Further analysis shows that we can interpret this filter as a Wiener filter on graphs. We finally build upon this interpretation to improve our results.

(Français) La ville et son double numérique

Sorry, this entry is only available in French.