About me
My research mainly focuses on developing scalable machine learning algorithms for networked data and assessing their accuracy performance both in theory and in practice.
I am a member of the LIFL lab and the MAGNET team (at INRIA Lille - NORD Europe), led by Prof. Marc Tommasi.
Since October 2013, I am an Assistant Professor (Maître de Conférences) at University of Lille (France).
I received an MSc degree (cum laude) in Computer Science from the University of Insubria and a PhD in Machine Learning / Computer Science from the Milan University (2011), working under the supervision of Prof. Nicolò Cesa-Bianchi and Prof. Claudio Gentile.
Previously, during my BSc and MSc, I spent about one year at the Computer Science Department of the University of Copenhagen, working with Prof. Jyrki Katajainen on designing novel and efficient data structures and assessing their practical performance.
In 2007 I worked with Prof. Claudio Gentile on developing and analyzing new sequential classification methods based on the Perceptron algorithm.
I have also some industrial experience: I worked at SAS Institute (Statistical Analysis System) in Italy (2007) and Bloomberg LP in Switzerland (2012/2013).
- N. Cesa-Bianchi, C. Gentile, F. Vitale, G. Zappella
Random Spanning Trees and the Prediction of Weighted Graphs
Journal of Machine Learning Research, 14:1251-1284, 2013. - N. Cesa-Bianchi, C. Gentile, F. Vitale
Predicting the labels of an unknown graph via adaptive exploration.
Theoretical Computer Science, 412(19):1791-1804, 2011, Special Issue on Algorithmic Learning Theory 2009. - J. Katajainen, F. Vitale
Navigation Piles with Applications to Sorting, Priority Queues, and Priority Deques.
Nordic Journal of Computing 10(3): 238- (2003).
(Mentioned by Donald E. Knuth in "The Art of Computer Programming" -- Section 7.1.3 - Bit tricks & techniques -- Vol. 4A).
Peer-reviewed conference publications
- M.Herbster, S. Pasteris, F. Vitale.
Online Sum-Product Computation over Trees
Proceedings of the 26th Conference on Neural Information Processing Systems -
NIPS 2012. - N. Cesa-Bianchi, C. Gentile, F. Vitale, and G. Zappella
A linear time active learning algorithm for link classification
Proceedings of the 26th Conference on Neural Information Processing Systems -
NIPS 2012.
(Extended version of "A Linear Time Active Learning Algorithm for Link Classification", 29th International Conference on Machine Learning - ICML 2012 - Workshop: Mining and Learning with Graphs). - N. Cesa-Bianchi, C. Gentile, F. Vitale, and G. Zappella
A correlation clustering approach to link classification in signed networks.
25th Annual Conference on Learning Theory. - COLT 2012 - JMLR Workshop and Conference Proceedings, 23:34.1-34.20, 2012. - F. Vitale, N. Cesa-Bianchi, C. Gentile, and G. Zappella
See the tree through the lines: the Shazoo algorithm
Proceedings of the 25th Annual Conference on Neural Information Processing Systems - NIPS 2011.
Full paper
- M. Herbster, S. Pasteris and F.Vitale
Efficient Prediction for Tree Markov Random Fields in a Streaming Model
25th Annual Conference on Neural Information Processing Systems - NIPS 2011 - Workshop: Discrete Optimization in Machine Learning (DISCML): Uncertainty, Generalization and Feedback. - N. Cesa-Bianchi, C. Gentile, F. Vitale, and G. Zappella
Active Learning on Graphs via Spanning Trees
24th Annual Conference on Neural Information Processing System - NIPS 2010 - Workshop: Networks Across Disciplines in Theory and Applications. - N. Cesa-Bianchi, C. Gentile, F. Vitale, and G. Zappella
Active Learning on Trees and Graphs
Proceedings of the 23rd Annual Conference on Learning Theory - COLT 2010. - N. Cesa-Bianchi, C. Gentile, F. Vitale, and G. Zappella
Random Spanning Trees and the Prediction of Weighted Graphs
Proceedings of the 27th International Conference on Machine Learning - ICML 2010. - N. Cesa-Bianchi, C. Gentile and F. Vitale
Fast and Optimal Algorithms for Weighted Graph Prediction
23rd Annual Conference on Neural Information Processing System - NIPS 2009 - Workshop: Analyzing Networks and Learning with Graphs. - N. Cesa-Bianchi, C. Gentile and F. Vitale
Learning Unknown Graphs
Proceedings of the 20th International Conference on Algorithmic Learning Theory - ALT 2009. - N. Cesa-Bianchi, C. Gentile and F. Vitale
Fast and Optimal Prediction on a Labeled Tree
Proceedings of the 22nd Annual Conference on Learning Theory - COLT 2009. - N. Cesa-Bianchi, C. Gentile and F. Vitale
Online Graph Prediction with Random Trees
22nd Annual Conference on Neural Information Processing System - NIPS 2008 - Workshop: New Challenges in Theoretical Machine Learning: Data Dependent Concept Spaces. - N. Cesa-Bianchi, C. Gentile and F. Vitale
On Higher-Order Perceptron Algorithms
Proceedings of the 21st Annual Conference on Neural Information Processing System - NIPS 2007.
Technical reports
- Claus Jensen, Jyrki Katajainen, and Fabio Vitale
Experimental Evaluation of Local Heaps
CPHSTL Report 2006-1 (2006). - Claus Jensen, Jyrki Katajainen, and Fabio Vitale
An Extended Truth about Heaps
CPH STL Report 2003-5 (2003).
Journal publications
Industrial experience
- Bloomberg LP, Switzerland (2012/2013).
- SAS Institute (Statistical Analysis System), Italy (2007).
Talks
- Copenhagen University (Denmark), 2013. Machine Learning on Trees and Graphs.
- Lille University & INRIA Lille - Nord Europe (France), 2013. Fast Prediction Algorithms for Networked Data.
- Microsoft Research, (Redmond, USA), 2012. Fast Learning on Graphs.
- Conference on Learning Theory (Edinburgh, Scotland) - COLT 2012. A Correlation Clustering Approach to Link Classification in Signed Networks.
- Conference on Neural Information Processing System (Granada, Spain) - NIPS 2011. See
the Tree through the Lines: the Shazoo Algorithm (spotlight presentation).
- International Conference on Machine Learning (Haifa, Israel) - ICML 2010. Random
Spanning Trees and the Prediction of Weighted Graphs.
- Conference on Learning Theory (Haifa, Israel) - COLT 2010. Active Learning on Trees
and Graphs.
- Conference on Algorithmic Learning Theory (Porto, Portugal) - ALT 2009. Learning
Unknown Graphs.
- Conference on Neural Information Processing System (Vancouver, Canada) - NIPS 2009 -
Workshop: Analyzing Networks and Learning with Graphs. Fast and Optimal
Algorithms for Weighted Graph Prediction.
- Conference on Learning Theory (Montreal, Canada) - COLT 2009. Fast and Optimal
Prediction on a Labeled Tree.
- Conference on Neural Information Processing System (Vancouver, Canada) - NIPS 2008 -
Workshop: New Challenges in Theoretical Machine Learning: Data Dependent
Concept Spaces. Online Graph Prediction with Random Trees.
- 4th STL-workshop Copenhagen 2003 (Denmark) . Navigation Piles with Applications to Sorting, Priority Queues, and Priority Deques.
Miscellaneous
- LIFL lab
(Laboratoire d'Informatique Fondamentale de Lille) - MAGNET team
(MAchine learninG in information NETworks). - PASCAL 2 Network of Excellence
EU, Seventh Programme Framework. - Data-dependent geometries and structures
("pump-priming project" within Pascal2). - Google-sponsored project SPAN
(Scalable Prediction Algorithms for Networked Data - Google Research Award, January 2010). - PASCAL Network of Excellence
EU, Sixth Programme Framework. - Generic programming - algorithms and tools
Danish Natural Science Research Council, 2005. (Project description). - Performance Engineering Laboratory
University of Copenhagen.
Research collaborators and colleagues
- Nicolò Cesa-Bianchi, University of Milan.
- Claudio Gentile, University of Insubria.
- Mark Herbster, University College London.
- Jyrki Katajainen, University of Copenhagen.
- Francesco Orabona, Senior Research Scientist at Yahoo! Labs NY.
- Stephen Pasteris, University College London.
- Marc Tommasi, Lille University & INRIA Lille.
- Giovanni Zappella, Machine Learning Scientist at Amazon Development Center Germany.
Visits to universities and research institutes
- May 2014: University of Milan, Department of Computer Science (Italy).
- November 2013: University of Milan, Department of Computer Science (Italy);
University College London, Department of Computer Science (United Kingdom);
University of Copenhagen, Department of Computer Science (Denmark); - July 2013: Lille University & INRIA Lille - Nord Europe, Department of Computer Science (France).
- July 28th - August 4th 2012: Microsoft Research, (Redmond, USA).
- November 2010; September 2011; November 2011; February 2012, May 2012: University College London, Department of Computer Science (United Kingdom).
- November 2008: University of Bonn, Department of Computer Science (Germany).
- September 2002: University of Copenhagen, Department of Computer Science (Denmark).
Teaching and tutoring
- Information Coding
University of Lille (Undergraduate level course - Academic Year 2014-2015, Spring semester). - Unsupervised Classification
University of Lille (Graduate level course - Academic Year 2014-2015, Spring semester). - Unsupervised Classification
University of Lille (Undergraduate level course - Academic Year 2014-2015, Spring semester). - Algorithms for Graphs
University of Lille (Undergraduate level course - Academic Year 2014-2015, Fall semester). - Unsupervised Classification
University of Lille (Graduate level course - Academic Year 2013-2014, Spring semester). - Artificial Intelligence
University of Lille (Undergraduate level course - Academic Year 2013-2014, Spring semester). - Information Technology and Internet Certificate ("Certificat informatique et Internet")
University of Lille (Undergraduate level course - Academic Year 2013-2014, Spring semester). - Information and Communication Technology for Education ("Technologies de l'Information et de la Communication pour l'Enseignement")
University of Lille (Undergraduate level course - Academic Year 2013- 2014, Spring semester). - Statistical Methods for Machine Learning
University of Milan (Graduate level course - Academic Year 2009-2010 Spring semester) - Java Programming Laboratory
University of Milan (Undergraduate level course - Academic Year 2009-2010 Fall semester) - Java Programming Laboratory
University of Milan (Undergraduate level course - Academic Year 2008-2009 Fall semester)
PhD students
- Geraud Le Falher, INRIA Lille - Nord Europe, 2014 (co-supervision).
Thesis supervision
- Giovanni Zappella, MSc in Computer Science, Università degli Studi di Milano - 2010 (role: co-supervisor).
International schools
- Bertinoro international Spring School, (Bertinoro, Italy), 2009.
- Bristol Summer School on Probabilistic Techniques in Computer Science, University of Bristol (United Kingdom), 2008.
- International PhD School on Randomized Algorithms (BiCi-SNS), Scuola Normale di Pisa (Italy), 2008.
- Summer School on Experimental Algorithmics, (Rungsted Kyst, Denmark), 2004.