CAp 2017 - Conférence sur l'Apprentissage Automatique

Program

Please respect the duration of your presentation

[L#] Long : 20 minutes followed by 10 minutes of questions
[C#] Short : 10 minutes followed by 5 minutes of questions
[P#] Poster : 5 minutes
Papers of the competition: TBA
Articles ATLAS : TBA

	Wednesday, June the 28th	Thursday, June the 29th	Friday, June the 30th
	Auditorium IMAG	Auditorium IMAG	Auditorium IMAG
8:30	Welcome
9:00	Invited speaker Yurii Nesterov, UCL	Invited speaker Léon Bottou, Facebook	Invited speaker Yves Grandvalet, CNRS
9:30
10:00
	Coffee Break
10:30	Session 1 L1, L2, L3, C1, C2	Session 3 L7, L8, L9, C5, C6	Session 4 L10, L11, L12, C7, C8
11:00
11:30
12:00
12:30
	Buffet


14:00	Session 2 L4, L5, L6, C3, C4	Session Competition	General Assembly
14:30			General Assembly
15:00			Session ATLAS
15:30
16:00
16:30	Presentation Poster 1	Presentation Poster 2
17h00	Session Posters (Wine, Cheese & Beer!)
17:30
18:00

20:00		Banquet Restaurant le téléphérique



24h00

Wednesday 28/06

Invited speaker - 9h - 10h

Complexity bounds for primal-dual methods minimizing the model of objective function
Yurii Nesterov, UCL

Session 1 - 10h30 - 12h30

Chair: Franck Iutzeler

[L1] Apprentissage collaboratif et décentralisé de modèles personnels sur un réseau
Paul Vanhaesebrouck, Aurélien Bellet, Marc Tommasi

[L2] Snake: a Stochastic Proximal Gradient Algorithm for Structured Regularized Problems over Large Graphs
Adil Salim, Pascal Bianchi, Walid Hachem

[L3] Gated End-to-End Memory Networks
Julien Perez, Fei Liu

[C1] Symbolic Priors for RNN-based Semantic Parsing
Chunyang Xiao, Marc Dymetman and Claire Gardent

[C2] Multi-view Generative Adversarial Networks
Mickael Chen and Ludovic Denoyer

Session 2 - 14h00 - 16h00

Chair: Charlotte Laclau

[L4] Theoretical Analysis of Domain Adaptation with Optimal Transport
Ievgen Redko, Amaury Habrard, Marc Sebban

[L5] Collaborative clustering based on Algorithmic Information Theory
Pierre-Alexandre Murena, Jérémie Sublime, Basarab Matei, Antoine Cornuéjols

[L6] Optimisation du top rank avec le gradient boosting pour la détection d'anomalies
Jordan Frery, Marc Sebban, Olivier Caelen, Liyun He Guelton, Amaury Habrard

[C3] Un borne PAC-Bayésienne en espérance et son extension à l'apprentissage multivues
Anil Goyal, Emilie Morvant, Pascal Germain

[C4] Apprentissage pour la segmentation de la prostate en IRM: architectures multi-échelle résiduelles
Bruno Sciolla, Matthieu Martin, Rémi Souchon, Olivier Rouvière, Philippe Delachartre

Presentation Poster 1 - 16h00 - 16h30

Chair: Massih-Reza Amini

[P1] Un protocole décentralisé et robuste pour le calcul de moyenne sous contrainte de vie privée
Pierre Dellenbach, Jan Ramon, Aurélien Bellet

[P2] Scikit-SpLearn: a toolbox for the spectral learning of weighted automata compatible with scikit-learn
Denis Arrivault, Dominique Benielli, François Denis, Rémi Eyraud

[P3] Feature extraction with regularized siamese networks for outlier detection: application to epilepsy lesion detection
Zaruhi Alaverdyan, Carole Lartizien

[P4] Transfer Learning by boosting projections from target to source
Antoine Cornuéjols, Sema Akkoyunlu, Pierre-Alexandre Murena, Raphaël Olivier

[P5] Multi-scale streaming anomalies detection for time series
Kiran Bangalore Ravi

Session Poster 1 - 16h30 - 18h30

[L1], [L2], [L3], [L4], [L5], [L6], [C1], [C2], [C3], [C4], [P1], [P2], [P3], [P4], [P5]

Thursday 29/06

Invited speaker - 9h - 10h

Quels signaux pour l’apprentissage non-supervisé ?
Léon Bottou, Facebook

Session 3 - 10h30 - 12h30

Chair: Marianne Clausel

[L7] Une Approche à Base de Copules pour Modéliser les Dépendances entre les Thèmes dans un Flux de Documents
Hesam Amoualian, Marianne Clausel and Eric Gaussier

[L8] Identifier les utilisateurs d'un compte partagé en exploitant le moment de consommation
Clément Lesaege, Francois Schnitzler, Anne Lambert and Jean-Ronan Vigouroux

[L9] Survival Regression with Sequencing of Censored Clinical Data
Jianfei Zhang, Aurélien Bach, Shengrui Wang, Lifei Chen, Josiane Courteau and Alain Vanasse

[C5] Apprentissage multi-tâche pour l'analyse de sentiments dans les tweets
Simon Moura and Georgios Balikas

[C6] Structured Prediction by minimization of a Fisher Surrogate Loss
Mousaab Djerrab, Alexandre Garcia, Maxime Sangnier and Florence D'Alché-buc

Session Competition - 14h00 - 16h00

Chair: Cédric Lopez

Opening Presentation
Presentation of CAp 2017 challenge: NER for French Tweets (15 min.)

Keynote Speaker: Xavier Carreras (45 min.)
The Challenge of Extracting New Entities in Text: A Low-rank Matrix Learning Approach

Synapse at CAp 2017 NER challenge: Fasttext CRF (20 min + 10 min QAs)
Damien Sileo, Camille Pradel, Philippe Muller and Tim Van de Cruys

FNER-BGRU-CRF at CAp 2017 NER challenge: Bidirectional GRU-CRF for French Named Entity Recognition in Tweets (20 min + 10 min QAs)
Mourad Gridach, Hatem Haddad and Hala Mulki

Presentation Poster 2 - 16h00 - 16h30

Chair: Massih-Reza Amini

[P6] Binary seeds for generative adversarial networks
Arnaud Sors, Stéphane Bonnet, Laurent Vercueil, Jean-François Payen

[P7] Dialog state tracking, a machine reading approach using Memory Network
Julien Perez, Fei Liu

[P8] Hyperparameter Optimization of Deep Neural Networks: Combining Hyperband with Bayesian Model Selection
Hadrien Bertrand, Roberto Ardon, Matthieu Perrot, Isabelle Bloch.

[P9] Modèles heuristiques de classification des opinions à partir d'une étude comparative des réseaux profonds
Lisa Medrouk, Anna Pappa

[P10] Joint Structured Learning and Predictions under Logical Constraints in Conditional Random Fields
Jean-Luc Meunier

Session Poster 2 - 16h30 - 18h30

[L7], [L8], [L9], [L10], [L11], [L12], [C5], [C6], [C7], [C8], [P6], [P7], [P8], [P9], [P10]

Friday 30/06

Invited speaker - 9h - 10h

Évolution de l'apprentissage automatique et de sa perception : quel impact pour notre communauté scientifique ?
Yves Grandvalet, CNRS

Session 4 - 10h30 - 12h30

Chair: Ievgen Redko

[L10] Apprentissage de sphères maximales d'exclusion avec garanties théoriques
Guillaume Metzler, Brahim Belkasmi, Xavier Badiche, Stéphane Canu, Elisa Fromont, Amaury Habrard and Marc Sebban

[L11] Multi-View Metric Learning in Vector-Valued Kernel Spaces
Riikka Huusari, Hachem Kadri and Cécile Capponi

[L12] Random Fourier Features For Operator-Valued Kernels
Romain Brault, Markus Heinonen and Florence D'Alché-buc

[C7] L3SVMs : Machines à Vecteurs de Support Locales et Linéaires basées sur des Landmarks
Valentina Zantedeschi, Rémi Emonet and Marc Sebban

[C8] Options Discovery with Budgeted Reinforcement Learning
Aurélia Léon and Ludovic Denoyer

Session ATLAS - 14h30 - 16h30

Chair: Myriam Tami

TBA