Program

Monday, 29 October 2012

• 8:00 AM – 8:45 AM — Registration of the participants
  ROOM: Building ERATO, Hall.
• 8:45 AM – 9:15 AM — Word of welcome
  ROOM: Building ERATO, Grand Amphithéâtre.
• 9:15 AM – 11:00 AM — Tutorial 1 (ALT): Model selection theory – a tutorial with applications to learning (P. Massart)
  This tutorial can be attended by DS participants.
  ROOM: Building ERATO, Grand Amphithéâtre, SESSION CHAIR: Peter Grünwald.
• 11:00 AM – 11:30 AM — BREAK
  Building ERATO, Hall
• 11:30 AM – 12:30 PM — ALT and DS sessions held in parallel
  ALT Session 1: Relations between Models and Data
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Pascal Massart.
  • 11:30 AM – 12:00 PM. Peter Grünwald.
    The Safe Bayesian: Learning the Learning Rate via the Mixability Gap
  • 12:00 PM – 12:30 PM. Lev Reyzin.
    Data Stability in Clustering: A Closer Look
  DS Session 1: Generic Pattern Mining Techniques
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Hadzic Fedgja.
  • 11:30 AM – 12:00 PM. Rui Li and Stefan Kramer.
    Efficient Redundancy Reduced Subgroup Discovery via Quadratic Programming
  • 12:00 PM – 12:30 PM. Esther Galbrun and Angelika Kimmig.
    Towards Finding Relational Redescriptions
• 12:30 PM – 02:00 PM — LUNCH / STEERING COMMITTEE MEETING
  Building ERATO, 1st floor
• 02:00 PM – 03:45 PM — Tutorial 2 (DS): Exploring sequential data (G. Ritschard)
  This tutorial can be attended by ALT participants.
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Atsuyoshi Nakamura.
• 03:45 PM – 04:15 PM — BREAK
  Building ERATO, Hall
• 04:15 PM – 06:15 PM — ALT and DS sessions held in parallel
  ALT Session 2: Inductive Inference
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Nader Bshouty.
  • 04:15 PM – 04:45 PM. Sanjay Jain, Timo Kötzing and Frank Stephan.
    Enlarging Learnable Classes
  • 04:45 PM – 05:15 PM. Ziyuan Gao and Frank Stephan.
    Confident and Consistent Partial Learning of Recursive Functions
  • 05:15 PM – 05:45 PM. Sanjay Jain and Efim Kinber.
    Automatic Learning from Positive Data and Negative Counterexamples
  • 05:45 PM – 6:15 PM. Christophe Costa Florêncio and Sicco Verwer.
    Regular Inference as Vertex Coloring
  DS Session 2: Clustering
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Bruno Crémilleux.
  • 04:15 PM – 04:45 PM. Bi-Ru Dai and Chih-Heng Chung.
    LF-CARS: a Loose Fragment-based Consensus Clustering Algorithm with a Robust Similarity
  • 04:45 PM – 05:15 PM. Bruno M. Nogueira, Alipio M. Jorge and Solange Rezende.
    HCAC: semi-supervised hierarchical clustering using confidence-based active learning
  • 05:15 PM – 05:45 PM. Khoa Nguyen and Sanjay Chawla.
    Large Scale Spectral Clustering Using Resistance Distance and Spielman-Teng Solvers
• 06:30 PM — COCKTAIL
  Building ERATO, 1st floor

Tuesday, 30 October 2012

• 09:00 AM – 10:00 AM — Invited speaker I (ALT-DS): Declarative modeling for machine learning and data mining (L. De Raedt)
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Djamel Zighed.
• 10:00 AM – 10:30 AM — BREAK
  Building ERATO, Hall
• 10:30 AM – 12:00 PM — ALT and DS sessions held in parallel
  ALT Session 3: Bandit Problems
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Gilles Stoltz.
  • 10:30 AM – 11:00 AM. Emilie Kaufmann, Nathaniel Korda and Rémi Munos.
    Thompson Sampling: An Optimal Finite Time Analysis
  • 11:00 AM – 11:30 AM. Ronald Ortner, Daniil Ryabko, Peter Auer and Rémi Munos.
    Regret Bounds for Restless Markov Bandits
  • 11:30 AM – 12:00 PM. Alexandra Carpentier and Rémi Munos.
    Minimax Number of Strata for Online Stratified Sampling given Noisy Samples
  DS Session 3: Stream and Sequence Mining
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Marc Plantevit.
  • 10:30 AM – 11:00 AM. Kazumi Saito, Kouzou Ohara, Masahiro Kimura and Hiroshi Motoda.
    Burst Detection in a Sequence of Tweets based on Information Diffusion Model
  • 11:00 AM – 11:30 AM. Mickaël Fabrègue, Agnès Braud, Sandra Bringay, Florence Le Ber and Maguelonne Teisseire.
    Including spatial relations and scales within sequential pattern extraction
  • 11:30 AM – 12:00 PM. Carlos Ferreira, Joao Gama, Vítor Santos Costa, Vladimiro Miranda and Audun Botterud.
    Predicting Ramp Events with a Stream-based HMM framework
• 12:15 AM – 01:30 PM — LUNCH
  Building ERATO, 1st floor
• 01:30 PM – 02:00 PM — Business Meeting
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Thomas Zeugmann.
• 02:00 PM – 03:00 PM — Invited speaker II (ALT): Learnability Beyond Uniform Convergence (S. Shalev-Shwartz)
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Nicolas Vayatis.
• 3:00 PM – 4:30 PM — ALT and DS sessions held in parallel
  ALT Session 4: Statistical Learning Theory and Classification
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Shai Shalev-Shwartz.
  • 03:00 PM – 03:30 PM. Mehryar Mohri and Andres Muñoz Medina.
    New Analysis and Algorithm for Learning with Drifting Distributions
  • 03:30 PM – 04:00 PM. Shai Ben-David and Ruth Urner.
    On the Hardness of Covariate Shift Learning (and the Utility of Unlabeled Target Samples)
  • 04:00 PM – 04:30 PM. Hal Daumé III, Jeff M. Phillips, Avishek Saha, and Suresh Venkatasubramanian.
    Efficient protocols for Distributed Classification and Optimization
  DS Session 4: Applications in Emerging Domains
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Cécile Favre.
  • 03:00 PM – 03:30 PM. Dragan Gamberger, Drazen Lucanin and Tomislav Smuc.
    Descriptive modeling of systemic banking crises
  • 03:30 PM – 04:00 PM. Einoshin Suzuki, Emi Matsumoto and Asuki Kouno.
    Data Squashing for HSV Subimages by an Autonomous Mobile Robot
  • 04:00 PM – 04:30 PM. Novita Ikasari and Fedja Hadzic.
    An Assessment on Loan Performance from Combined Quantitative and Qualitative Data in XML
• 04:30 PM – 05:00 PM — BREAK
  Building ERATO, Hall
• 05:00 PM – 07:00 PM — ALT and DS sessions held in parallel
  ALT Session 5.1: Teaching and PAC-Learning
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Sanjay Jain.
  • 05:00 PM – 05:30 PM. Rahim Samei, Pavel Semukhin, Boting Yang and Sandra Zilles.
    Sauer's Bound for a Notion of Teaching Complexity
  • 05:30 PM – 06:00 PM. Dana Angluin, James Aspnes and Aryeh Kontorovich.
    On the Learnability of Shuffle Ideals
  ALT Session 5.2: Online Prediction of Individual Sequences (I)
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Wouter M. Koolen.
  • 06:00 PM – 06:30 PM. Edward Moroshko and Koby Crammer.
    Weighted Last-Step Min-Max Algorithm with Improved Sub-Logarithmic Regret
  • 06:30 PM – 07:00 PM. Daiki Suehiro, Kohei Hatano, Shuji Kijima, Eiji Takimoto and Kiyohito Nagano.
    Online Prediction under Submodular Constraints
  DS Session 5: Algorithmic Issues
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Ganberger Dragan.
  • 05:00 PM – 05:30 PM. Shunsuke Makino, Takaharu Shimada, Kouichi Hirata, Kouki Yonezawa and Kimihito Ito.
    A Trim Distance between Positions in Nucleotide Sequences
  • 05:30 PM – 06:00 PM. Koji Tabata, Atsuyoshi Nakamura and Mineichi Kudo.
    Fast Approximation Algorithm for the 1-Median Problem
  • 06:00 PM – 06:30 PM. Tom de Ruijter, Evgeni Tsivtsivadze and Tom Heskes.
    Online Co-Regularized Algorithms
• 07:30 PM — CONFERENCE BANQUET
  Restaurant Le Caro de Lyon, 25 rue du Bât d'argent

Wednesday, 31 October 2012

• 09:00 AM – 10:00 AM — Invited speaker III (DS): Recent Developments in Pattern Mining (T. Calders)
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Jean-Gabriel Ganacia.
• 10:00 AM – 10:30 AM — Winners of the Carl Smith Award (DS)
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Einoshin Suzuki.
  • 10:00 AM – 10:30 AM. Willy Ugarte, Patrice Boizumault, Samir Loudni and Bruno Crémilleux.
    Soft Threshold Constraints for Pattern Mining
• 10:30 AM – 11:00 AM — BREAK
  Building ERATO, Hall
• 11:00 AM – 12:30 PM — ALT and DS sessions held in parallel
  ALT Session 6: Online Prediction of Individual Sequences (II)
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Ronald Ortner.
  • 11:00 AM – 11:30 AM. Eyal Gofer and Yishay Mansour.
    Lower Bounds on Individual Sequence Regret
  • 11:30 AM – 12:00 PM. Dmitry Adamskiy, Wouter M. Koolen, Alexey Chernov and Vladimir Vovk.
    A Closer Look at Adaptive Regret
  • 12:00 PM – 12:30 PM. Gábor Bartók and Csaba Szepesvári.
    Partial monitoring with side information
  DS Session 6: Graph-Based Mining
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Motoda Hiroshi.
  • 11:00 AM – 11:30 AM. Elise Desmier, Marc Plantevit, Céline Robardet and Jean-François Boulicaut.
    Cohesive Co-Evolution Patterns in Dynamic Attributed Graphs
  • 11:30 AM – 12:00 PM. Kafil Hajlaoui, Pascal Cuxac, Jean-Charles Lamirel and Claire François.
    Enhancing Patent Expertise through Automatic Matching with Scientific Papers
  • 12:00 PM – 12:30 PM. Yoshiaki Okubo, Makoto Haraguchi and Etsuji Tomita.
    Structural Change Pattern Mining Based on Constrained Maximal k-Plex Search
• 12:30 PM – 02:00 PM — LUNCH
  Building ERATO, Hall
• 02:00 PM – 04:00 PM — ALT and DS sessions held in parallel
  ALT Session 7: Other Models of Online Learning
  ROOM: Building CLIO, Amphithéâtre Laprade, SESSION CHAIR: Eiji Takimoto.
  • 02:00 PM – 02:30 PM. Tor Lattimore and Marcus Hutter.
    PAC Bounds for Discounted MDPs
  • 02:30 PM – 03:00 PM. Wouter M. Koolen and Vladimir Vovk.
    Buy Low, Sell High
  • 03:00 PM – 03:30 PM. Manfred Warmuth, Wojciech Kotlowski and Shuisheng Zhou.
    Kernelization of Matrix Updates, When and How?
  • 03:30 PM – 04:00 PM. Satyadev Nandakumar and Mrinalkanti Ghosh.
    Predictive Complexity and Generalized Entropy Rate of Stationary Ergodic Processes
  DS Session 7: Machine Learning
  ROOM: Building ATHENA, Amphithéâtre Fugier, SESSION CHAIR: Julien Velcin.
  • 02:00 PM – 02:30 PM. Pierre Alquier and Xiaoyin Li.
    Prediction of quantiles by statistical learning and application to GDP forecasting
  • 02:30 PM – 03:00 PM. Johannes Fürnkranz and Sang-Hyeun Park.
    Error-Correcting Output Codes as a Transformation from Multi-Class to Multi-Label Prediction
  • 03:00 PM – 03:30 PM. Prem Raj Adhikari and Jaakko Hollmén.
    Fast Progressive Training of Mixture Models for Model Selection
  • 03:30 PM – 04:00 PM. Francis Maes, Raphael Fonteneau, Louis Wehenkel and Damien Ernst.
    Policy Search in a Space of Simple Closed-form Formulas: Towards Interpretability of Reinforcement Learning
• 04:00 PM – FINAL BREAK AND GOODBYE
  Building ERATO, Hall