Emilie Morvant
               Assistant Professor (Maître de Conférences) in Machine Learning

Machine Learning Transfer Learning, Domain Adaptation PAC-Bayes Bounds, Methods (Semi-)Supervised Learning Robustness, Adversarial learning Representation Learning Similarity Functions/Metric Learning Deep Learning Multiview/Multimodality Learning Fairness and biais in ML Highly unbalanced data (anomaly/fraud detection)

Research Projects Projet inter-laboratoire FIL (Fédération Informatique de Lyon) (with G. Metzler) Random Fourrier Features for PAC-Bayesian Domain Adaptation ANR project TAUDoS - member Theory and Algorithms for the Understanding of Deep learning On Sequential data ANR project APRIORI - coordinator A Pac-bayesian RepresentatIOn leaRnIng perspective INS2I-CNRS project JCJC 2018 "PaRaFF" (with P. Germain) PAC-Bayesian Random Fourier Features ANR project LIVES - member Learning with Interacting ViEwS My postdoc was funded by the ERC grant no308036: Life-Long Learning for Visual Scene Understanding I was a member of the PASCAL2 Network of excellence My Ph.D. was funded by VideoSense ANR project: Rich concepts recognition in multilingual videos, throw pivot languages I regularly participated in Lampada ANR Project: Learning Algorithms, Models and sPArse representations for structured DAta

---

Supervision

PostDoc

Marie-Ange Lebre --- Deep Learning for detection and classification of microorganism (since September 2022, co-supervised with A. Habrard and R. Emonet)

Ph.D. Student

Currently no PhD student

Former Ph.D. Students

Paul Viallard --- Beyond PAC-Bayesian Bounds: From Disintegration to Novel Bounds (from September 2019 to December 2022, co-supervised with A. Habrard and P. Germain) --- Now Post-doc at Inria Paris

Léo Gautheron --- Learning Tailored Data Representations from Few Labeled Examples (from October 2017 to December 2020, co-supervised with Pr. A. Habrard and Pr. M. Sebban) --- Now Data Science Engineer at Synapse Défense

Anil Goyal --- Learning a Multiview Weighted Majority Vote Classifier: Using PAC-Bayesian Theory and Boosting (from November 2015 to October 2018, co-supervised with Pr. Massih-Reza Amini) --- Now Lead Data Scientist at Housing.com (India)

Interns

Julien Bastian --- Random Fourier Features, PAC-Bayes and Domain Adaptation (Mar-July 2023, co-supervised with G. Metzler)
Alexiane Fraisse --- Random Fourier Features and Domain Adaptation (Apr-July 2022, co-supervised with G. Metzler and P. Viallard)
Luiza Dzhidzhavadze --- A Multiclass C-Bound-Based Algorithm (Apr-June 2021, co-supervised with P. Viallard)
Himanshu Pandey --- A Multiclass C-Bound-Based Algorithm (Apr-June 2021, co-supervised with P. Viallard)
Paul Viallard --- Deep Learning and PAC-Bayes (Feb-June 2019, co-supervised with A. Habrard and R. Emonet)
Omar El-Sabrout --- Active Learning for PAC-Bayesian Domain Adaptation (Mid-April Mid-July 2018)
Loujain Liekah --- Experts Combination (April-June 2018, co-supervised with Dr. M. Soare)
Luc Giffon --- Efficient anomaly detection in data stream (Feb-June 2017, co-supervised with Dr. A. Bonnefoy and Dr. T. Peel)
Arunava Maulik --- (April-June 2017, co-supervised with Pr. A. Habrard and Dr. M. Soare)
Prem Prakash --- (April-June 2017, co-supervised with Pr. A. Habrard)
Léo Gautheron --- Improving the bibliometry platform Labmetry (April-June 2016, co-supervised with Pr. M. Sebban)
Benjamin Sabot --- Empirical study of the C-bound as stopping criterion for neural networks (April-June 2016, co-supervised with Pr. A. Habrard, Dr. P. Germain and D. Fourure)
Soroush Seifi --- A PAC-Bayesian Multiview Study (April-June 2016, co-supervised with Pr. A. Habrard and A. Goyal)

PhD Jurys

2022 Luxin Zhang, Lille (external examinator)
2020 Léo Gautheron, St-Etienne (co-supervisor)
2018 Anil Goyal, St-Etienne/Grenoble (co-supervisor)

---

Teaching

Depuis Juillet 2015, je suis la responsable de la 2ème année de la licence d'informatique.
Since July 2015, I am in charge of the 2nd year of BSc in Computer Science.

S1

L2 - Systèmes d'exploitation
L2 - Programmation impérative (classique et spécifique alternants)

S2

L2 - Programmation impérative

Old teaching here - Anciens enseignements ici

---

Bibliography    (My DBLP webpage - My Google Scholar webpage)

 •  Ph.D. Thesis  •  Book  •  International Journals  •  International Conferences  •  International Workshops  •  French Conferences  •  Reports  •  Misc.  • 

Ph.D. Thesis (in french)

• Apprentissage de vote de majorité pour la classification supervisée et l’adaptation de domaine :
  approches PAC-Bayésiennes et combinaison de similarités
  Learning Majority Vote for Supervised Classification and Domain Adaptation:
  PAC-Bayesian Approaches and Similarity Combination
  PhD committee: Stéphane Ayache (co-advisor), Antoine Cornuéjols (chair, examiner), Rémi Gilleron (examiner), Amaury Habrard (advisor), Mario Marchand (reviewer), Liva Ralaivola (examiner), Michèle Sebag (reviewer).
  Defense date: September 18, 2013.
  Ph.D. award 2013 from Aix-Marseille University
  A runner-up award of the "Prix de thèse de l'AFIA 2014"
  [pdf] [bibtex] [slides] [abstract/related works]

Book

• Domain Adaptation Theory: Available Theoretical Results
  Ievgen Redko ; Emilie Morvant ; Amaury Habrard ; Marc Sebban ; Younès Bennani
  ISTE Press-Elsevier, 2019, ISBN : 9781785482366
  [publisher's website] [bibtex] [short version - ArXiV]

International Journals

• A General Framework for the Practical Disintegration of PAC-Bayesian Bounds
  Paul Viallard ; Pascal Germain ; Amaury Habrard ; Emilie Morvant
  Machine Learning Journal (MLJ), 2023, (accepted, to be published and presented at ECML-PKDD 2023)
  [research report arXiv:2102.08649]
  
  
• Metric Learning from Imbalanced Data with Generalization Guarantees
  Léo Gautheron ; Amaury Habrard ; Emilie Morvant ; Marc Sebban
  Pattern Recognition Letters (PRL), 133:298-304, 2020, DOI : 10.1016/j.patrec.2020.03.008
  [published version] [bibtex]
  
  
• PAC-Bayes and Domain Adaptation
  Pascal Germain ; Amaury Habrard ; François Laviolette ; Emilie Morvant
  Neurocomputing, 379:379-397, 2020, DOI : 10.1016/j.neucom.2019.10.105
  [pdf] [published version] [bibtex]
  
  
• Multiview Boosting by Controlling the Diversity and the Accuracy of View-specific Voters
  Anil Goyal ; Emilie Morvant ; Pascal Germain ; Massih-Reza Amini
  Neurocomputing, 358:81-92, 2019, DOI : 10.1016/j.neucom.2019.04.072
  [pdf] [published version][bibtex]
  
  
• Risk Upper Bounds for General Ensemble Methods with an application to Multiclass Classification
  François Laviolette, Emilie Morvant, Liva Ralaivola, Jean-Francis Roy
  Neurocomputing, 219:15-25, 2017, DOI: 10.1016/j.neucom.2016.09.016
  [pdf] [published version] [bibtex]
  
  
• Domain Adaptation of Weighted Majority Votes via Perturbed Variation-Based Self-Labeling
  Emilie Morvant
  Pattern Recognition Letters (PRL), 51(0):37–43, 2015, DOI: 10.1016/j.patrec.2014.08.013
  [pdf] [published version] [bibtex]
  
  
• Learning A Priori Constrained Weighted Majority Votes
  Aurélien Bellet ; Amaury Habrard ; Emilie Morvant ; Marc Sebban
  Machine Learning Journal (MLJ), 97(1-2):129-154, 2014, DOI: 10.1007/s10994-014-5462-z
  [pdf] [published version] [bibtex]
  
  
• Parsimonious Unsupervised and Semi-Supervised Domain Adaptation with Good Similarity Functions
  Emilie Morvant ; Amaury Habrard ; Stéphane Ayache
  Knowledge and Information Systems (KAIS), 33(2):309-349, 2012, DOI: 10.1007/s10115-012-0516-7
  [pdf] [published version] [bibtex]

International Conferences

• Leveraging PAC-Bayes Theory and Gibbs Distributions for Generalization Bounds with Complexity Measures
  Paul Viallard ; Rémi Emonet ; Amaury Habrard ; Emilie Morvant ; Valentina Zantedeschi
  nternational Conference on Artificial Intelligence and Statistics (AISTATS), 2024
  [pdf]
  
  
• Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound
  Valentina Zantedeschi ; Paul Viallard ; Emilie Morvant ; Rémi Emonet ; Amaury Habrard ; Pascal Germain ; Benjamin Guedj
  Conference on Neural Information Processing Systems (NeurIPS), 2021
  [pdf]
  
  
• A PAC-Bayes Analysis of Adversarial Robustness
  Guillaume Vidot ; Paul Viallard ; Amaury Habrard ; Emilie Morvant
  Conference on Neural Information Processing Systems (NeurIPS), 2021
  [pdf]
  
  
• Self-Bounding Majority Vote Learning Algorithms by the Direct Minimization of a Tight PAC-Bayesian C-Bound
  Paul Viallard ; Pascal Germain ; Amaury Habrard ; Emilie Morvant
  European Conference on Machine Learning & Principles and Pratice of Knowledge Discovery in Databases (ECML-PKDD), 2021
  [pdf] [bibtex]
  
  
• Landmark-based Ensemble Learning with Random Fourier Features and Gradient Boosting
  Léo Gautheron ; Pascal Germain ; Amaury Habrard ; Guillaume Metzler ; Emilie Morvant ; Marc Sebban ; Valentina Zantedeschi
  European Conference on Machine Learning & Principles and Pratice of Knowledge Discovery in Databases (ECML-PKDD), 2020
  [pdf] [bibtex]
  
  
• Metric Learning from Imbalanced Data
  Léo Gautheron ; Amaury Habrard ; Emilie Morvant ; Marc Sebban
  The IEEE International Conference on Tools with Artificial Intelligence (ICTAI), 2019, Portland, Oregon, USA
  [pdf] [bibtex]
  
  
• Pseudo-Bayesian Learning with Kernel Fourier Transform as Prior
  Gaël Letarte ; Emilie Morvant ; Pascal Germain
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2019, Naha, Okinawa, Japan
  [pdf][bibtex][research report arXiv:1810.12683]
  
  
• Multiview Learning of Weighted Majority Vote by Bregman Divergence Minimization
  Anil Goyal ; Emilie Morvant ; Massih-Reza Amini
  International Symposium on Intelligent Data Analysis (IDA), 2018, 's-Hertogenbosch, the Netherlands
  [pdf][bibtex]
  
  
• PAC-Bayesian Analysis for a two-step Hierarchical Mutliview Learning Approach
  Anil Goyal ; Emilie Morvant ; Pascal Germain ; Massih-Reza Amini
  European Conference on Machine Learning & Principles and Pratice of Knowledge Discovery in Databases (ECML-PKDD), 2017, Skopje, Macedonia
  [pdf][bibtex][research report arXiv:1606.07240]
  
  
• A New PAC-Bayesian Perspective on Domain Adaptation
  Pascal Germain ; Amaury Habrard ; François Laviolette ; Emilie Morvant
  International Conference on Machine Learning (ICML), 2016, New York, USA
  [pdf][bibtex][research report arXiv:1506.04573] [video] [DALC code] [extended journal version]
  
  
• Multilabel Structured Output Learning with Random Spanning Trees of Max-Margin Markov Networks
  Mario Marchand ; Su Hongyu ; Emilie Morvant ; Juho Rousu ; John Shawe-Taylor
  Neural Information Processing Systems (NIPS), 2014, Montréal, Canada
  [pdf] [bibtex]
  
  
• A PAC-Bayesian Approach for Domain Adaptation with Specialization to Linear Classifiers
  Pascal Germain ; Amaury Habrard ; François Laviolette ; Emilie Morvant
  International Conference on Machine Learning (ICML), 2013, Atlanta, USA
  [pdf] [bibtex] [video] [PBDA code] [extended version]
  
  
• The Multi-Task Learning View of Multimodal Data
  Hachem Kadri ; Stéphane Ayache ; Cécile Capponi ; Sokol Koço ; François-Xavier Dupé ; Emilie Morvant
  Asian Conference on Machine Learning (ACML), 2013, Canberra, Australia
  [pdf] [bibtex]
  
  
• PAC-Bayesian Generalization Bound on Confusion Matrix for Multi-Class Classification
  Emilie Morvant ; Sokol Koço ; Liva Ralaivola
  International Conference on Machine Learning (ICML), 2012, Edinburgh, United Kingdom. pp. 815-822
  [pdf] [bibtex] [video] [research report arXiv:1202.6228]
  
  
• Sparse Domain Adaptation in Projection Spaces based on Good Similarity Functions
  Emilie Morvant ; Amaury Habrard; Stéphane Ayache
  IEEE International Conference on Data Mining series (ICDM), 2011, Vancouver, Canada. IEEE Computer Society, pp. 457-466
  Selected as one of the best papers for possible publication in Knowledge and Information Systems (KAIS).
  [pdf] [bibtex]

International Workshops

Published

• Majority Vote of Diverse Classifiers for Late Fusion
  Emilie Morvant ; Amaury Habrard ; Stéphane Ayache
  IAPR Joint International Workshops on Statistical Techniques in Pattern Recognition (SPR) and Structural and Syntactic Pattern Recognition (SSPR), 2014, Joensuu, Finland. pp. 153-162
  [pdf] [bibtex] [research report arXiv:1207.1019]
  
  
• On the Usefulness of Similarity Based Projection Spaces for Transfer Learning
  Emilie Morvant ; Amaury Habrard ; Stéphane Ayache
  First International Workshop on Similarity-Based Pattern Recognition, 2011, Venise, Italy. pp. 1-16
  Organized by the EU FP7 Project SIMBAD
  [pdf] [bibtex] [video]

Benchmarks

• VideoSense at TRECVID 2011: Semantic Indexing from Light Similarity Functions-based Domain Adaptation with Stacking
  Emilie Morvant; Stéphane Ayache; Amaury Habrard; Miriam Redi; Tanase Claudiu; Bernard Merialdo; Bahjat Safadi; Franck Thollard; Nadia Derbas; Georges Quenot
  TRECVID Workshop participants notebook papers, NIST, 2012
  [pdf] [bibtex]

Not Published

• Interpreting Neural Networks as Majority Votes through the PAC-Bayesian Theory
  Paul Viallard ; Rémi Emonet ; Pascal Germain ; Amaury Habrard ; Emilie Morvant
  NeurIPS 2019 Workshop on Machine Learning with guarantees, Vancouver, Canada.
  [pdf]
  
  
• A New PAC-Bayesian View of Domain Adaptation
  Pascal Germain ; François Laviolette ; Amaury Habrard ; Emilie Morvant
  NIPS 2015 Workshop on Transfer and Multi-Task Learning: Trends and New Perspectives, Montréal, Canada.
  [research report arXiv:1506.04573] [DALC code]
  Finalized version published at ICML'16 [bibtex]
  
  
• On Generalizing the C-Bound to the Multiclass and Multi-label Settings
  François Laviolette ; Emilie Morvant ; Liva Ralaivola ; Jean-Francis Roy
  NIPS 2014 Workshop on Representation and Learning Methods for Complex Outputs , Montréal, Canada.
  [pdf] [research report arXiv:1408.1336]
  
  
• An Improvement to the Domain Adaptation Bound in a PAC-Bayesian context
  Pascal Germain ; François Laviolette ; Amaury Habrard ; Emilie Morvant
  NIPS 2014 Workshop on Transfer and Multi-task learning: Theory Meets Practice, Montréal, Canada.
  [pdf] [research report arXiv:1503.06944]
  
  
• Domain Adaptation of Majority Votes via Perturbed Variation-based Label transfer
  Emilie Morvant
  NIPS 2013 Workshop New Directions in Transfer and Multi-Task: Learning Across Domains and Tasks, Lake Tahoe, Nevada, United States.
  [pdf]
  Finalized version published in PRL [bibtex]
  
  
• PAC-Bayesian Approach for Domain Adaptation with Specialization to Linear Classifiers
  Emilie Morvant (Pascal Germain ; Amaury Habrard ; François Laviolette)
  Women in Machine Learning workshop (WiML 2013), Lake Tahoe, Nevada, United States.
  Presentation of our ICML'13 paper
  
  
• PAC-Bayesian Learning and Domain Adaptation
  Pascal Germain ; Amaury Habrard ; François Laviolette ; Emilie Morvant
  NIPS 2012 Workshop Multi-trade-off in Machine Learning, Lake Tahoe, Nevada, United States.
  [pdf] [poster]
  Finalized version published at ICML'13 [bibtex] [video]
  
  
• Generalization of the C-bound to Multiclass Setting
  Emilie Morvant (François Laviolette ; Liva Ralaivola ; Jean-Francis Roy)
  Women in Machine Learning workshop (WiML 2012), Lake Tahoe, Nevada, United States.
  
  
• Sparse Domain Adaptation in a Good Similarity-Based Projection Space
  Emilie Morvant ; Amaury Habrard ; Stéphane Ayache
  NIPS 2011 Domain Adaptation Workshop, Sierra Nevada, Spain.
  [pdf] [poster]
  Finalized version published at ICDM'11 [bibtex]

French Conferences

• Intérêt des bornes désintégrées pour la généralisation avec des mesures de complexité
  Paul Viallard ; Rémi Emonet ; Pascal Germain ; Amaury Habrard ; Emilie Morvant ; Valentina Zantedeschi
  French Conference on Machine Learning (CAp 2022), 2022
  
  
• Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound
  Valentina Zantedeschi ; Paul Viallard ; Emilie Morvant ; Rémi Emonet ; Amaury Habrard ; Pascal Germain ; Benjamin Guedj
  French Conference on Machine Learning (CAp 2022), 2022
  [pdf, published at NeurIPS 21]
  
  
• Apprentissage de Vote de Majorité par Minimisation d’une C-Borne
  Paul Viallard ; Emilie Morvant ; Pascal Germain
  French Conference on Machine Learning (CAp 2021), 2021
  
  
• Dérandomisation des Bornes PAC-Bayésiennes
  Paul Viallard ; Emilie Morvant ; Pascal Germain
  French Conference on Machine Learning (CAp 2021), 2021
  
  
• Une Analyse PAC-Bayésienne de la Robustesse Adversariale
  Guillaume Vidot ; Paul Viallard ; Emilie Morvant
  French Conference on Machine Learning (CAp 2021), 2021
  
  
• Théorie PAC-Bayésienne pour l'apprentissage en deux étapes de réseaux de neurones
  Paul Viallard ; Rémi Emonet ; Amaury Habrard ; Emilie Morvant; Pascal Germain
  French Conference on Machine Learning (CAp 2020), 2020
  
  
• Apprentissage d'ensemble basé sur des points de repère avec des caractéristiques de Fourier aléatoires et un renforcement du gradient
  Léo Gautheron ; Pascal Germain ; Amaury Habrard ; Guillaume Metzler ; Emilie Morvant ; Marc Sebban ; Valentina Zantedeschi French Conference on Machine Learning (CAp 2020), 2020
  
  
• Revisite des "random Fourier features" basée sur l'apprentissage PAC-Bayésien via des points d'intérêts
  Léo Gautheron, Pascal Germain, Amaury Habrard, Gaël Letarte, Emilie Morvant, Marc Sebban, Valentina Zantedeschi
  French Conference on Machine Learning (CAp 2019), 2019, Toulouse, France.
  
  
• Apprentissage d’un vote de majorité hiérarchique pour l’apprentissage multivue
  Anil Goyal, Emilie Morvant, Massih-Reza Amini
  French Conference on Machine Learning (CAp 2018), 2018, Rouen, France.
  
  
• Apprentissage de métrique pour la classification supervisée de données déséquilibrées
  Léo Gautheron ; Amaury Habrard ; Emilie Morvant ; Marc Sebban
  French Conference on Machine Learning (CAp 2018), 2018, Rouen, France.
  
  
• Une borne PAC-Bayésienne en espérance et son extension à l'apprentissage multivues
  Anil Goyal, Emilie Morvant, Pascal Germain
  French Conference on Machine Learning (CAp 2017), 2017, Grenoble, France.
  
  
• Théorèmes PAC-Bayésiens pour l'apprentissage multi-vues
  Anil Goyal, Emilie Morvant, Pascal Germain, Massih-Reza Amini
  French Conference on Machine Learning (CAp 2016), 2016, Marseille, France.
  
  
• Adaptation de domaine de vote de majorité par auto-étiquetage non itératif
  Emilie Morvant
  French Conference on Machine Learning (CAp 2014), 2014, Saint-Etienne, France. pp. 49-58
  [pdf]
  English version published in PRL [bibtex]
  
  
• Une analyse PAC-Bayésienne de l’adaptation de domaine et sa spécialisation aux classifieurs linéaires
  Pascal Germain ; Amaury Habrard ; François Laviolette ; Emilie Morvant
  French Conference on Machine Learning (CAp 2013), 2013, Lille, France.
  [pdf] [PBDA code]
  English version published at ICML'13 [bibtex]
  
  
• Vote de majorité a priori contraint pour la classification binaire : spécification au cas des plus proches voisins.
  Aurélien Bellet ; Amaury Habrard ; Emilie Morvant ; Marc Sebban
  French Conference on Machine Learning (CAp 2013), 2013, Lille, France.
  [pdf]
  English version published in MLJ [bibtex]
  
  
• Étude de la généralisation de DASF à l'adaptation de domaine semi-supervisée
  Emilie Morvant ; Amaury Habrard ; Stéphane Ayache
  French Conference on Machine Learning (CAp 2012), 2012, Nancy, France. pp. 111-126
  [pdf]
  English version published in KAIS [bibtex]
  
  
• Adaptation de domaine parcimonieuse par pondération de bonnes fonctions de similarité
  Emilie Morvant ; Stéphane Ayache ; Amaury Habrard
  French Conference on Machine Learning (CAp 2011), 2011, Chambéry, France. pp. 295-310
  [pdf]
  English version published at ICDM'11 [bibtex]

Research Reports --- not (yet) published

• Semi-Universal Adversarial Perturbations
  Jordan Frecon ; Paul Viallard ; Emilie Morvant ; Gilles Gasso ; Amaury Habrard ; Stéphane Canu
  [research report]
  
  
• Generalization Bounds with Arbitrary Complexity Measures
  Paul Viallard ; Rémi Emonet ; Amaury Habrard ; Emilie Morvant ; Valentina Zantedeschi
  [research report]
  
  
• PAC-Bayesian Theorems for Domain Adaptation with Specialization to Linear Classifiers (2015)
  Pascal Germain ; Amaury Habrard ; François Laviolette ; Emilie Morvant
  [research report arXiv:1503.06944]
  Extended version of our ICML'13 paper

Miscellaneous

• Proceedings of The 38th Annual Workshop of the Austrian Association for Pattern Recognition (ÖAGM), 2014
  Vladimir Kolmogorov ; Christoph Lampert ; Emilie Morvant ; Rustem Takhanov
  [proceedings] [website of the conference]
  
  
• Flexible Domain Adaptation for Multimedia Indexing (Best Poster Award, presentation of a part of my ICDM'11 paper)
  Emilie Morvant ; Amaury Habrard ; Stéphane Ayache
  Poster presentation at ENS/INRIA Computer Vision and Machine Learning Summer School, 2011.
  [poster]