Marc Tommasi
- Machine Learning in Information Networks (MAGNET)
Présentation
Professeur en informatique à l'université de Lille. Je suis membre du laboratoire CRIStAL . Je dirige l'équipe Magnet.. Depuis 2015, mes recherches portent sur l'apprentissage automatique et les réseaux d'information, ainsi que sur l'apprentissage décentralisé et l'apprentissage automatique préservant la confidentialité. Auparavant, j'ai travaillé principalement sur les automates d'arbres et les contraintes entre 1990 et 2000, puis sur la classification, l'apprentissage automatique et l'extraction de données et d'informations structurées (2000-2015).
Publications
Publications
|
|
Tau-Eval: A Unified Evaluation Framework for Useful and Private Text AnonymizationProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, Nov 2025, Suzhou, France. pp.216-227, ⟨10.18653/v1/2025.emnlp-demos.16⟩ |
|
|
TAMIS: Tailored Membership Inference Attacks on Synthetic DataECML PKDD 2025 - European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Sep 2025, Porto (Portugal), Portugal. pp.203-220, ⟨10.1007/978-3-032-06096-9_12⟩ |
|
|
Exploiting Context-dependent Duration Features for Voice Anonymization Attack SystemsInterspeech 2025, Aug 2025, Rotterdam, Netherlands |
|
|
Privacy Amplification Through Synthetic Data: Insights from Linear RegressionICML 2025 - 42nd International Conference on Machine Learning, 2025, Vancouver, Canada. ⟨10.48550/arXiv.2506.05101⟩ |
|
|
Analysis of Speech Temporal Dynamics in the Context of Speaker Verification and Voice Anonymization2025 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025), Apr 2025, Hyderabad, India. ⟨10.1109/ICASSP49660.2025.10887896⟩ |
|
|
TAROT: Task-Oriented Authorship Obfuscation Using Policy Optimization MethodsProceedings of the Sixth Workshop on Privacy in Natural Language Processing, Apr 2025, Albuquerque, United States. pp.14-31, ⟨10.18653/v1/2025.privatenlp-main.2⟩ |
|
|
Rényi Pufferfish Privacy: General Additive Noise Mechanisms and Privacy Amplification by Iteration via Shift Reduction LemmasInternational Conference on Machine Learning (ICML 2024), 2024, Vienna (Austria), Austria |
|
|
Improved Stability and Generalization Guarantees of the Decentralized SGD AlgorithmICML 2024 - The Forty-first International Conference on Machine Learning, Jul 2024, Vienne, Austria |
|
|
Refined Convergence and Topology Learning for Decentralized SGD with Heterogeneous DataProceedings of The 26th International Conference on Artificial Intelligence and Statistics (AISTATS 2023), 2023, Valencia, Spain, Spain |
|
|
High-Dimensional Private Empirical Risk Minimization by Greedy Coordinate DescentAISTATS 2023 - International Conference on Artificial Intelligence and Statistics, Apr 2023, Valencia, Spain. pp.4894-4916 |
|
|
Differential Privacy has Bounded Impact on Fairness in ClassificationInternational Conference on Machine Learning, Jul 2023, Honolulu, United States |
|
|
Extraction d'informations liées au locuteur depuis un modèle acoustique personnaliséJEP 2022, Jun 2022, île de Noirmoutier, France |
|
|
FLamby: Datasets and Benchmarks for Cross-Silo Federated Learning in Realistic Healthcare SettingsNeurIPS 2022 - Thirty-sixth Conference on Neural Information Processing Systems, Nov 2022, New Orleans, United States |
|
|
Differentially Private Coordinate Descent for Composite Empirical Risk MinimizationICML 2022 - 39th International Conference on Machine Learning, Jul 2022, Baltimore, United States. pp.14948-14978 |
|
|
Enhancing speech privacy with slicingInterspeech 2022 - Human and Humanizing Speech Technology, Sep 2022, Incheon, South Korea |
|
|
Retrieving Speaker Information from Personalized Acoustic Models for Speech RecognitionIEEE ICASSP 2022, 2022, Singapour, Singapore |
|
|
On speaker verification from the neural network footprint of personalized acoustic modelsJournées d'Études sur la Parole - JEP2022, Jun 2022, Île de Noirmoutier, France |
|
|
Privacy attacks for automatic speech recognition acoustic models in a federated learning frameworkICASSP 2022, 2022, Singapour, Singapore |
|
|
Study on Acoustic Model Personalization in a Context of Collaborative Learning Constrained by Privacy PreservationSPECOM 2021 - 23rd International Conference on Speech and Computer, Sep 2021, St Petersburg, Russia. pp.426 - 436, ⟨10.1007/978-3-030-87802-3_39⟩ |
|
|
A comparative study of speech anonymization metricsINTERSPEECH 2020, Oct 2020, Shanghai, China |
|
|
Design Choices for X-vector Based Speaker AnonymizationINTERSPEECH 2020, International Speech Communication Association (ISCA), Oct 2020, Shanghai, China |
|
|
Evaluating Voice Conversion-based Privacy Protection against Informed AttackersICASSP 2020 - 45th International Conference on Acoustics, Speech, and Signal Processing, IEEE Signal Processing Society, May 2020, Barcelona, Spain. pp.2802-2806 |
|
|
Fully Decentralized Joint Learning of Personalized Models and Collaboration GraphsAISTATS 2020 - The 23rd International Conference on Artificial Intelligence and Statistics, Aug 2020, Palerme / Virtual, Italy |
|
|
Privacy-Preserving Adversarial Representation Learning in ASR: Reality or Illusion?INTERSPEECH 2019 - 20th Annual Conference of the International Speech Communication Association, Sep 2019, Graz, Austria |
Communication-Efficient Decentralized Boosting while Discovering the Collaboration GraphNeurIPS 2018 - Workshop on Machine Learning on the Phone and other Consumer Devices, 2nd Workshop on Machine Learning on the Phone and other Consumer Devices (MLPCD 2), Dec 2018, Montréal, Canada |
|
|
|
Personalized and Private Peer-to-Peer Machine LearningAISTATS 2018 - 21st International Conference on Artificial Intelligence and Statistics, Apr 2018, Lanzarote, Spain. pp.1-20 |
|
|
Decentralized Collaborative Learning of Personalized Models over NetworksInternational Conference on Artificial Intelligence and Statistics (AISTATS), Apr 2017, Fort Lauderdale, Florida., United States |
Personalized and Private Peer-to-Peer Machine LearningNIPS 2017 - Workshop on Machine Learning on the Phone and other Consumer Devices, Dec 2017, Long Beach, United States |
|
|
|
Hypernode Graphs for Spectral Learning on Binary Relations over SetsECML/PKDD - 7th European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Sep 2014, Nancy, France |
|
|
Learning Sequential Tree-to-Word Transducers8th International Conference on Language and Automata Theory and Applications, Mar 2014, Madrid, Spain |
|
|
Hypernode Graphs for Spectral Learning on Binary Relations over SetsConférence Francophone sur l'Apprentissage Automatique (Cap 2014), Jul 2014, Saint-Etienne, France. ⟨10.1007/978-3-662-44851-9_42⟩ |
|
|
Clustering Spectral avec Contraintes de Paires réglées par Noyaux GaussiensCAP 2014, Jul 2014, Saint-Etienne, France |
|
|
Fast Gaussian Pairwise Constrained Spectral ClusteringECML/PKDD - 7th European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Sep 2014, Nancy, France. pp.242 - 257, ⟨10.1007/978-3-662-44848-9_16⟩ |
|
|
Learning from Multiple Graphs using a Sigmoid KernelThe 12th International Conference on Machine Learning and Applications (ICMLA'13), Dec 2013, Miami, United States |
|
|
Spectral Estimation of Conditional Random Graph Models for Large-Scale Network dataUAI 2012 - 28th Conference on Uncertainty in Artificial Intelligence, 2012, Avalon, United States |
|
|
Fiedler Random Fields: A Large-Scale Spectral Approach to Statistical Network ModelingNeural Information Processing Systems (NIPS), Dec 2012, Lake Tahoe, United States |
|
|
Normalization of Sequential Top-Down Tree-to-Word Transducers5th International Conference on Language Automata Theory and Appliciations, May 2011, Tarragona, Spain |
|
|
Relevant Representations for the Inference of Rational Stochastic Tree LanguagesInternational Colloquium on Grammatical Inference, 2008, St Malo, France. pp.57-70 |
Automatic Wrapper Induction from Hidden-Web Sources with Domain KnowledgeInternational Workshop on Web information and data managment, Oct 2008, Napa, United States. pp.9-16 |
|
|
|
On Probability Distributions for Trees: Representations, Inference and LearningNIPS Workshop on Representations and Inference on Probability Distributions, Dec 2007, Whistler, Canada |
|
|
Conditional Random Fields for XML TreesWorkshop on Mining and Learning in Graphs, Sep 2006, Berlin, Germany |
|
|
Champs Conditionnels Aléatoires pour l'Annotation d'ArbresCAp 2006, May 2006, Trégastel, France |
|
|
Interactive Tuples Extraction from Semi-Structured DataWeb Intelligence, Dec 2006, Hong Kong, China |
XML Document Transformation with Conditional Random FieldsINEX 2006, Dec 2006, Dagstuhl, Germany |
|
|
|
Querying Unranked Trees with Stepwise Tree Automata19th International Conference on Rewriting Techniques and Applications, 2004, Aachen, Georgia. pp.105--118 |
|
|
Learning Multi-label Alternating Decision Trees from Texts and DataInternational Conference on Machine Learning and Data Mining, 2003, Leipzig, Georgia. pp.35-49 |
|
|
Residual Finite Tree Automata7th International Conference on Developments in Language Theory, Jul 2003, Szeged, Hungary, Hungary. pp.171 -- 182 |
Text Classification from Positive and Unlabeled ExamplesProceedings of the 9th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems, IPMU'02, 2002, Grenoble, France. pp.1927--1934 |
|
Homomorphisms and concurrent term rewritingFCT 1999 - International Symposium on Fundamentals of Computation Theory, Aug 1999, Iasi, Romania. pp.475-487, ⟨10.1007/3-540-48321-7_40⟩ |
|
Deciding the Satisfiability of Quantifier free Formulae on One-Step RewritingRewriting Techniques and Applications, 10th International Conference, RTA-99, Trento, Italy, July 2-4, 1999, Proceedings, 1999, Unknown, Unknown Region. pp.103--117, ⟨10.1007/3-540-48685-2_9⟩ |
|
Some new Decidability Results on Positive and Negative Set ConstraintsProceedings of First International Conference on Constraints in Computational Logics, CCL'94, 1994, Munich, Germany. pp.336--351 |
|
Solving Systems of Set Constraints using Tree AutomataProceedings of the 10th Annual Symposium on Theoretical Aspects of Computer Science, STACS'93, 1993, Würzburg, Germany. pp.505--514 |
|
Solving Systems of Set Constraints with Negated Subset RelationshipsProceedings of the $34^th$ Symposium on Foundations of Computer Science, FOCS'93, 1993, Palo Alto California, United States. pp.372--380 |
|
|
Hypernode Graphs for Learning from Binary Relations between Groups in NetworksNetworks: From Graphs to Rich Data, NIPS Workshop., Dec 2014, Montreal, Canada. , 2014 |
|
|
Tree Automata Techniques and Applicationspp.262, 2008 |
Champs Markoviens Conditionnels pour l'extraction d'informationEric Gaussier and François Yvon. Modèles probabilistes pour l'accès à l'information textuelle, Hermès, pp.223-267, 2011 |
Sequential Tree-to-Word Transducers: Normalization, Minimization, and Learning2015 |
|
L’apprentissage automatique : le diable n’est pas dans l’algorithme2015 |
|
|
Learning with Locally Private Examples by Inverse Weierstrass Private Stochastic Gradient Descent2026 |
|
|
Federated Learning for MRI-based BrainAGE: a multicenter study on post-stroke functional outcome prediction2025 |
|
|
Probability Estimation over Large-Scale Random Networks via the Fiedler Delta Statistic2013 |
|
|
Structures arborescentes et apprentissage automatiqueAutre [cs.OH]. Université Charles de Gaulle - Lille III, 2006 |