MLG 2008 banner

Programme

The workshop is located at the University of Helsinki main building at Fabianinkatu 33. (Google map).

All talks as well as the poster spotlights will be held in the Small hall (4th floor) of the main building, except the second part of Session 4, starting at 16:00 on Saturday, will be held in hall S5 (3rd floor). The poster session will take place in the 2nd floor lobby of the main building.

The workshop registration and information desk is located (together with ICML/COLT/UAI) at the entrance lobby of the main building, and will be open throughout the workshop.


Friday, July 4

8:00-Registration
8:45-9:00Welcome (Small hall, 4th floor)

Session 1

Chair: Imre Risi Kondor

9:00-10:00Invited speaker: Four graph partitioning algorithms
Fan Chung, University of California, San Diego
10:00-10:30Combining near-optimal feature selection with gSpan.
Karsten Borgwardt, Xifeng Yan, Marisa Thoma, Hong Cheng, Arthur Gretton, Le Song, Alex Smola, Jiawei Han, Philip Yu and Hans-Peter Kriegel.
10:30-11:00Coffee break
11:00-11:30Representative Subgraph Sampling using Markov Chain Monte Carlo Methods
Christian Hübler, Karsten Borgwardt, Hans-Peter Kriegel and Zoubin Ghahramani
11:30-12:00Inferring the structure and scale of modular networks
Jake M. Hofman and Chris H. Wiggins
12:00-12:30Min, Max and PTIME Anti-Monotonic Overlap Graph Measures
Toon Calders, Jan Ramon and Dries Van Dyck
12:30-13:30Lunch Break

Session 2

Chair: Jyrki Kivinen

13:30-14:30Invited speaker: Influence and Correlation in Social Networks
Mohammad Mahdian, Yahoo! Research
14:30-15:00Classification in Graphs using Discriminative Random Walks
Jérôme Callut, Kevin Françisse, Marco Saerens and Pierre Dupont
15:00-15:30An Online Algorithm for Learning a Labeling of a Graph
Kristiaan Pelckmans and Johan A.K. Suykens
15:30-16:00Coffee break
16:00-16:30A New Kernel for Classification of Networked Entities
Dell Zhang and Robert Mao
16:30-17:00Induction of Node Label Controlled Graph Grammar Rules
Hendrik Blockeel and Siegfried Nijssen
17:00-17:30Poster spotlights
(Chair: Hannes Heikinheimo)
17:30-19:30Poster session
20:30-Banquet

Saturday, July 5

Session 3

Chair: Craig Saunders

9:00-10:00Invited speaker: Structured Output Prediction with Structural SVMs
Thorsten Joachims, Cornell University
10:00-10:30Efficient Discriminative Training Method for Structured Predictions
Huizhen Yu, Dimitri P. Bertsekas and Juho Rousu
10:30-11:00Coffee break
11:00-11:30Comparing Diffusion Models for Graph–Based Semi–Supervised Learning
Aram Galstyan and Paul R. Cohen
11:30-12:00Structure and tie strengths in a mobile communication network
Jari Saramäki and Jukka-Pekka Onnela
12:00-12:30Improved Software Fault Detection with Graph Mining
Frank Eichinger, Klemens Böhm and Matthias Huber
12:30-14:00Lunch Break

Session 4

Chair: Luc De Raedt

14:00-15:00Invited speaker: Biomine search engine for probabilistic graphs
Hannu Toivonen, University of Helsinki
15:00-15:30Parameter Learning in Probabilistic Databases: A Least Squares Approach
Bernd Gutmann, Angelika Kimmig, Luc De Raedt and Kristian Kersting
15:30-16:00Coffee break
(The session continues in hall S5!)
16:00-16:30Infinite mixtures for multi-relational categorical data
Janne Sinkkonen, Janne Aukia and Samuel Kaski
16:30-17:00Markov Logic Improves Protein β-Partners Prediction
Marco Lippi and Paolo Frasconi
17:00-17:30A Hilbert-Schmidt Dependence Maximization Approach to Unsupervised Structure Discovery
Matthew B. Blaschko and Arthur Gretton

17:30Closing