Academic Team

Florence d'Alché-Buc, the holder of the Data Science & Artificial Intelligence for Digitalized Industry & Services Chair, has been since 2014 a full professor at Télécom Paris, an IMT grande école. Previously, she was a professor at Université d’Evry, an ATIGE (Genopole Thematic Incentive Actions) researcher and joint head of the IBISC lab. She launched and managed the Challenges program as part of the PASCAL European network (2004-08) and since 2017, has become the scientific director of the Digiscome Labex. Her research is on machine learning, network inference, structured prediction and dynamical system modeling. She has authored more than 80 articles in international journals and conference proceedings.

As part of her teaching work, she is co-head of the Data Science Master 2 at Université Paris-Saclay jointly awarded by École Polytechnique, ENSAE Paris and Université Paris Sud. She participated in the conception of the new continuous education programs in artificial intelligence: a Post-Master Degree and a specialist certificate. She is also responsible for the Bearing Point Data Science Education Chair.

Keywords: machine learning, kernel methods, structured prediction, network inference, dynamical systems.

Visit her website

Chloé Clavel is a lecturer and professor at Télécom Paris, where she facilitates the Social Computing group (https://www.tsi.telecom-paristech.fr/en/research/1885-2/social-computing-topic/). She is currently working on the interaction between humans and machines, including the analysis of users’ socio-emotional behavior and strategies for socio-emotional interaction. She has contributed to several European and national collaborative projects relating to Social Computing (e.g. H2020 ITN ANIMATAS, aria-valuspa UE-TIC, Labex smart).

Keywords: opinion analysis, natural language processing, learning, speech processing, emotion recognition, human-machine interaction.

Visit her website

Stephan Clémençon is a full professor at Télécom Paris, Institut Mines-Télécom, and head of the S2A Research Team (Signal, Statistics and Learning). He carries out his research activity in applied math in the Télécom Paris LTCI Lab. His research topics are mainly related to machine learning, probability and statistics. He is in charge of the Big Data Post-Master Degree at Télécom Paris and held the Machine Learning for Big Data Chair from 2013 to 2018.

Keywords: ranking, clustering, anomaly detection, graph-mining, recommendation systems

Visit his website

Stéphane Lathuilière is an associate professor (maître de conférences) at Télécom Paris, Institut Polytechnique de Paris, France, in the LTCI's Multimedia team. Until October 2019, he was a post-doctoral fellow at the University of Trento in the Multimedia and Human Understanding Group, led by Prof. Nicu Sebe and Prof. Elisa Ricci. He received the M.Sc. degree in applied mathematics and computer science from ENSIMAG, Grenoble Institute of Technology (Grenoble INP), France, in 2014. He completed his master thesis in the International Research Institute MICA (Hanoi, Vietnam). He worked towards his Ph.D. in mathematics and computer science in the Perception Team at Inria under the supervision of Dr. Radu Horaud, and obtained it from Université Grenoble Alpes (France) in 2018.

His research interests cover machine learning for computer vision problems (eg. domain adaptation, continual learning) and deep models for image and video generation. He published papers in the most prestigious computer vision conferences (CVPR, ICCV, ECCV, NeurIPS) and top journals (T-PAMI).

Keywords: computer vision, domain adaptation, continual learning, deep models for image and video generation

Visit his website

Pavlo Mozharovskyi joined Télécom Paris as an Assistant Professor in 2018. After having finished his studies at Kyiv Polytechnic Institute in automation control and informatics, he obtained a PhD degree at the University of Cologne in 2014, where he conducted research in nonparametric and computational statistics and classification. He has been a postdoc at Agrocampus Ouest in Rennes with the Centre Henri Lebesgue for a year working on imputation of missing values, and then joined the CREST laboratory at the National School of Statistics and Information Analysis. His main research interests lie in the areas of statistical data depth function, classification, computational statistics, robust statistics, missing values, and data envelopment analysis.

Keywords: data depth, computational statistics, robust statistics, missing values, data envelopment analysis.

Visit his website

François Roueff is a professor at Télécom Paris, IMT, working in the S2A research team, and the deputy director of the Hadamard mathematical doctoral school. His research areas lie mainly in the fields of statistical signal processing and the analysis and the random modeling of time series and statistics for stochastic processes.

Keywords: long dependency, wave analysis, Hawkes process, locally stationary processes.

Visit his website

Roland Badeau is Full Professor in the Signal, Statistics and Machine learning (S2A) team of the Image, Data, Signal (IDS) Department at Télécom Paris. His research interests focus on statistical modeling of non-stationary signals (including adaptive high-resolution spectral analysis and Bayesian extensions to NMF), with applications to audio and music (source separation, denoising, dereverberation, multipitch estimation, automatic music transcription, audio coding, audio inpainting). He is a co-author of 30 journal papers, over 100 international conference papers, and 4 patents. He is also an Associate Editor of the EURASIP Journal on Audio, Speech, and Music Processing and the IEEE/ACM Transactions on Audio, Speech, and Language Processing.

Visit his website

Pascal Bianchi was born in 1977. He was awarded an MSc from Université Paris 11 and Supélec in 2000 and a doctorate from Université de Marne-la-Vallée in 2003. He was an assistant professor at the Supélec Telecommunication Department from 2003 to 2009. He then joined the Signal, Statistics and Learning (S2A) team at the Télécom Paris LTCI lab. His current research interests are stochastic optimization, probability and signal processing.

Keywords: statistical signal processing, convex optimization, distributed optimization, network sensors.

Visit his website

Albert Bifet is Full Professor at Télécom Paris. He is one of the leaders of MOA and Apache SAMOA software environments for implementing algorithms and running experiments for online learning from evolving data streams. He is co-author of a book on Machine Learning from Data Streams at MIT Press. Previously he worked at Huawei Noah's Ark Lab in Hong Kong, Yahoo Labs in Barcelona, University of Waikato and UPC BarcelonaTech. He was serving as Co-Chair of the Industrial track of IEEE MDM 2016, ECML PKDD 2015, and as Co-Chair of BigMine (2018-2012), and ACM SAC Data Streams Track (2019-2012).

Keywords: Artificial Intelligence, Data Streams, IoT, Real-Time Analytics, Machine Learning, Graph Mining.

Visit his website

Thomas Bonald is a professor at Télécom Paris and heads the DIG Team (Data, Intelligence and Graphs). His research interests are in the analysis of network-structured data. He is one of the main contributors of scikit-network, a Python package for the analysis of massive graphs. He is also interested in multivariate time-series and NLP.

Keywords: Large graphs, spectral methods, clustering, recommender systems.

Visit his website

Slim Essid is Full Professor at Télécom Paris and the coordinator of the Audio Data Analysis and Signal Processing theme (ADASP). His research is on machine learning and artificial intelligence for temporal data analysis, especially multiview learning, structured prediction, representation learning and data decomposition methods. The target applications include machine listening and music content analysis (MIR); multimodal perception: human behavior analysis, affective computing, and EEG data analysis; multimedia content analysis, especially joint audiovisual data analysis.

Keywords: Multiview learning, structured prediction, representation learning, audio and multimodal data.

Visit his website

Olivier Fercoq is a lecturer at Télécom Paris. He obtained a Master’s diploma from Université Paris 6 and ENSTA Paris. He studied optimization problems linked to web page references and applications in biology during his PhD at École polytechnique (2009-2012). He spent two years at the University of Edinburgh where he worked on coordinate descent methods. He joined Télécom ParisTech in 2014. His current research focuses on developing and studying optimization algorithms for high-dimensional problems.

Keywords: optimization, stochastic algorithms, rate of convergence, high dimensional, parallel computation.

Jhony Giraldo is Associate Professor at Télécom Paris, Institut Polytechnique de Paris, in the Multimédia team of the LTCI.

Jhony Giraldo received a Bachelor in Electronics Engineering and a Master of Science degree at Universidad de Antioquia, Colombia, in 2016 and 2018, respectively. He spent 15 months at the University of Delaware, USA, working on Graph Signal Processing between 2018 and 2019. Jhony was a visiting scholar at the Università degli Studi di Napoli Parthenope, Italy, in 2021. In 2022, he was a visiting Ph.D. student at CentraleSupélec, Université Paris-Saclay, France, before finishing his Ph.D. in Applied Mathematics at La Rochelle Université, France.

His research interests include the fundamentals and applications of Graph Neural Networks, Computer Vision, Machine Learning, and Graph Signal Processing.

Ekhine Irurozki received a Ph.D. in Computer Science from the University of the Basque Country, Spain, in 2014. Since 2020 she has been Assistant Professor at Télécom Paris. Before joining Télécom Paris, she was a postdoc fellow at the Machine Learning Department at the Basque Center for Applied Mathematics (BCAM) in Bilbao from 2017 to 2020. She previously worked from 2015 to 2017 as Postdoctoral Fellow at the Basque Country University. Her main research interest is the development of optimization and artificial intelligence techniques for automatic decision making, with a particular interest in statistical problems with ranked data.

Dr. Irurozki has led the cybersecurity lab and has been part of the transfer unit at BCAM, having worked in several projects with companies in the manufacturing and services sectors. She has received the Extraordinary Doctoral Award from the Basque Country University.

Yann Issartel is an associate professor at Télécom Paris, Institut Polytechnique de Paris. Until August 2022, he was a post-doctoral fellow at the Center for Research in Economics and Statistics (CREST) at ENSAE, Institut Polytechnique de Paris, under the supervision of Cristina Butucea. He obtained his M.S. in Probability and Statistics at the Institut de Mathématiques d'Orsay, Paris-Saclay University, and conducted, under the supervision of professors Christophe Giraud and Nicolas Verzelen, a PhD in mathematical statistics and learning up to 2020 at the same institution.

His research interests include data privacy and random networks. He has already published for conferences (NeurIPS) or prestigious scientific reviews (Journal of Machine Learning Research)

Keywords: random networks, latent position models, link prediction, sequential learning, bandits, data privacy, generative adversarial network

Ons Jelassi is a continuing education teacher at Télécom Paris and Télécom Evolution. She supervises Post Master’s Degree and Specialist Certificate programs in data science and artificial intelligence. She has worked in the fields of metrology and network performance as part of her doctoral research and as an audit and expertise consultant for large companies. Her work in the LTCI S2A research team is on the scalability of machine learning algorithms.

Keywords: machine learning, performance, scalability, distributed algorithms.

Matthieu Labeau joined Télécom Paris as a Senior Lecturer in 2019. He completed his doctorate at the University of Paris-Saclay, and became interested in the problems posed by large vocabularies in language modeling. He then became a postdoctoral fellow at the University of Edinburgh. His research areas, within automatic language processing, mainly concern representation learning and language modeling.

Keywords: automatic natural language processing, language modeling, word representations, deep learning

Charlotte Laclau is an Associate Professor at Télécom Paris, Institut Polytechnique de Paris since Sept. 2022 . She's a member of the LTCI lab, and part of the S2A team. Prior to that, she was holding the same position at Hubert Curien Laboratory in the Data Intelligence team. Her main line of research is in machine learning. Her research focus includes representation learning for complex data, fairness in machine learning for relational data and unsupervised learning for high-dimensional data.

Keywords: Machine learning, complex data, fairness

Laurence Likforman has been an associate professor (certified for doctoral supervision) at Télécom Paris since 1991. She heads the Pattern Recognition course and also teaches Signal Processing, Document Analysis and Statistics. Her research includes the use of recurrent and convolutional neural networks in the recognition and analysis of handwriting.

Keywords: handwriting recognition, Markov models, Bayesian networks, recurrent neural networks.

Visit her website

Geoffroy Peeters is full-professor in the LTCI S2A team at Télécom Paris. He received his PHDs degree in 2001 and Habilitation in 2013 from University Paris-VI on audio signal processing, data analysis and machine learning. Before joining Télécom Paris, he lead research related to Music Information Retrieval at IRCAM. His current research work is on signal processing, machine learning and deep learning applied to audio and music data analysis.

Keywords: audio signal processing, machine learning and deep learning.

Visit his website

Gaël Richard is Professor at Télécom Paris, Institut Mines-Télécom and head of the Image, Data, Signal (IDS) department. His research work lies at the core of digitization and is dedicated to the analysis, transformation, understanding and automatic indexing of acoustic signals (including speech, music, surrounding sounds) and to a lesser extent of heterogeneous and multimodal signals. In particular, he developed several source separation methods for audio and musical signals based on machine learning approaches.

Keywords: Machine listening, Matrix Factorization, Representation and subspace learning, Music Information Retrieval (MIR), Sound recognition, Audio source separation.

Visit his website

Fabian M. Suchanek is a full professor at Télécom Paris. He developed inter alia the YAGO knowledge base, one of the largest public general-purpose knowledge bases, which earned him a honorable mention of the SIGMOD dissertation award. His interests include information extraction, automated reasoning, and knowledge bases. Fabian has published around 70 scientific articles, among others at ISWC, VLDB, SIGMOD, WWW, CIKM, ICDE, and SIGIR, and his work has been cited more than 7000 times.

Keywords: semantic web, knowledge bases, information extraction, natural language processing, automatic reasoning.

Visit his website

Enzo Tartaglione received a joint Master’s degree in Electronic Engineering from: Politecnico di Torino, University of Illinois at Chicago and Politecnico di Milano in 2015. He received the Alta Scuola Politecnica diploma in 2016. In 2019, he defended his PhD thesis at Politecnico di Torino, on the topic “From Statistical Physics to Deep Neural Network Algorithms”. From 2019 to 2021, he held a postdoctoral position at the University of Turin, working on the European project  DeepHealth. His research topics include neural networks pruning, compression, quantization, regularization, deep learning applied to medical image processing, data and model debiasing, privacy preservation.

Keywords: neural networks pruning, compression, quantization, regularization, deep learning, model debiasing, privacy preservation

Visit his Google Scholar page

Anass Aghbalou is currently in the second year of his thesis at Telecom Paris. He obtained an engineering degree from École Centrale de Lyon in 2018 and a master's degree in statistical and financial engineering - Machine Learning track, from Université Paris-Dauphine in 2019. After a year spent as a research engineer at Telecom Paris, he is, since December 2020, a PhD student under the supervision of Anne Sabourin (Telecom Paris), François Portier (ENSAI Rennes) and Patrice Bertail (University of Paris Nanterre). His research focuses on resampling techniques, in particular cross-validation, for machine learning algorithms dedicated to rare events (anomaly detection, unbalanced classification, extreme values).

Dimitri Bouche is working on a topic that is little explored in machine learning: space-time processes. He will be analyzing functional and operator-valued kernel methods, with a view to meeting several of the challenges inherent to this type of data. Dimitri Bouche is working with Florence d’Alché-Buc, the head of the DSAIDIS Chair and a professor at Télécom Paris, François Roueff, professor at Télécom Paris and Marianne Clausel, professor at the University of Lorraine.

Tamim El Ahmad is in Year 2 of the Mathematics, Vision and Learning (MVA) Master’s program at ENS Paris-Saclay, a Master’s in mathematics, specialized in machine learning and computer vision.  He was previously in Year 1 of the Applied Mathematics Master’s program at Paris-Diderot and in the Master’s Degree in Engineering at Mines Saint-Etienne. He integrated the team on the 27/04/2020 for the following internship: “The best of both worlds: deep kernel  learning”, supervised by Florence d'Alché-Buc, professor at Télécom Paris and holder of the DSAIDIS Chair. His focus was on the development of hybrid architectures combining neural networks and kernel methods in order to solve prediction problems relating to structured data (especially link prediction, sequence and graph prediction). Since the end of his internship the 30/09/2020, he is doing a thesis, still under the supervision of Florence d'Alché-Buc.

Key words: deep learning, kernel methods, structured data prediction, hybrid structure

Marc Hulcelle is a first year PhD student at Télécom Paris since October 2020, under the supervision of Chloé Clavel (Télécom Paris) and Giovanna Varni (Télécom Paris). His research focuses on automatic analysis of social behaviors in human-robot interaction, with a special interest on trust and engagement. He completed his engineering degree from the école des Mines de Saint-Etienne. He worked on the robot Buddy at Blue Frog Robotics before starting his thesis.

Keywords: machine learning, human-robot interaction, affective computing.

Jayneel Parekh is a PhD student at Telecom Paris since September 2019. He is working under the guidance of Florence d'Alché-Buc and Pavlo Mozharovskyi. His thesis focuses on making machine learning models interpretable by design. He completed his bachelors and masters from IIT Bombay in electrical engineering and is primarily interested in machine learning and its applications to audio/image domains.

Keywords: Interpretable machine learning, deep learning, audio/image analysis

Emilia Siviero is a first year PhD student  at Télécom Paris. Her advisors are Stephan Clémençon (Télécom Paris) and Emilie Chautru (Mines ParisTech). Her research focuses on statistical learning methods for spatial data. She graduated from the Master 2 of Mathematics, Vision and Learning (MVA) of ENS Paris-Saclay.

Keys words: statistical learning, spatial data, geostatistics.

Jérémy Guérin works under the direction of Pavlo Mozharovskyi on the topic of data depth. My PhD will begin on April 2022 but I work on the subject since November 2021. This subject uses different courses I took during my studies : high dimensional statistics and optimisation. Data depth allows to quantify the centrality of a datapoint inside of a multivariate dataset. I try to prove the existence of algorithms able to compute the depth for datasets of large dimension so that they can be used in practice.

Keywords: data depth, machine learning, optimisation, robustness, computational statistics.

Nathan Huet is a PhD student since September 2021 under the supervision of Stephan Clémençon and Anne Sabourin (Télécom Paris). He first obtained a Bachelor's degree in Fundamental and Applied Mathematics, then obtained his Master's degree at the University of Paris-Saclay in Mathematics of Randomness, with a specialization in Probability and Statistics. After a research internship at Télécom Paris from 14/04/2021 to 30/09/2021, with the topic: "Extreme Value Analysis of Functional Data and Applications to the Detection of Functional Anomalies", under the supervision of Stephan Clémençon and Anne Sabourin, Nathan continued with a thesis, supervised by the same team. His research focuses on the processing of atypical functional data for anomaly detection. The probabilistic framework adopted is that of regular variation and extreme value theory in functional spaces.

Keywords: Extreme value theory, Functional analysis, Dimension reduction

Junjie Yang is in Year 2 of the École d’Ingénieurs ParisTech Shanghai Jiao Tong (SJTU-ParisTech) Master’s program. He has carried out research on automatic natural language processing, especially around such issues as question answering and natural language understanding via machine learning.

He will join the team from 18/05/2020 to 17/11/2020 for the following internship: “Output representation for the prediction of structured text”, supervised by Matthieu Labeau, associate professor at Télécom Paris and Florence d’Alché-Buc, professor at Télécom Paris and holder of the DSAIDIS Chair. The aim of the internship is to apply several groups of methods to predict the structured output of text data, in particular the generation of text, which may have to perform a large number of potentially very different tasks (abstractive summarization, question answering, automatic image description, machine translation…).

Key words: automatic natural language processing, structured output prediction, text generation, deep learning

Sanjeel Parekh will be focussing on active learning. He will be advised by Prof. Florence d'Alché-Buc. Previously, he was a postdoctoral research engineer working on real-time audio event detection for a startup project at Telecom Paris. From 2016-19, he worked towards his Ph.D. thesis on audio-visual scene analysis at Technicolor R&D and Telecom Paris, France. Prior to that, he received M.S. in Sound and Music Computing from Universitat Pompeu Fabra (UPF), Spain in 2015 and B. Tech (hons.) degree in electronics and communication engineering from LNM Institute of Information Technology, India in 2014. His research focuses on developing and applying machine learning techniques to problems in audio/visual domains.

Keywords: Representation learning, deep learning, active learning, audio-visual analysis

Arturo Castellanos is a research engineer in the team led by Florence d'Alché-Buc. Inside DSAIDIS chair, he was part of a project with Valeo in order to detect anomalies on production lines, and is now contributing to software library for anomaly detection. He graduated as engineer from Telecom Paris & EURECOM, and has a Msc in Computer Science from Oxford.

Keywords: anomaly detection, kernel methods.

Ignacio Laurenty is graduated from the Master 2 - Random Modeling, Finance and Data Science of Paris City University. He is currently working on active learning using bandits in the DSAIDIS chair, under the supervision of Florence d’Alché-Buc and Ekhine Irurozki. Using a hierarchy over labels, he studies the feasibility of learning policies to partially label data for classifications tasks.

Keywords: bandits, active learning, data label, classification

See his LinkedIn page

Sarah is currently working as a research engineer within the S2A team under the supervision of Florence d'Alché-Buc. Her research interests are mainly in Sequence Modeling and Representation Learning for time-varying graph-structured data, with a focus on economic applications. Sarah holds a Master's degree in Statistics and Econometrics from Toulouse School of Economics (TSE) and an Advanced Master's in Data Science from ENSAE Paris.

Keywords : Graph-structured data, Machine Learning, Representation learning, Structured sequence prediction, Time Series.

Asma Atamna is a postdoctoral researcher in the LTCI S2A team at Télécom Paris since September 2019. She works on Deep Learning approaches for analyzing multimodal signals in Human-Robot Interaction. She received her Ph.D. in Black-Box Continuous Optimization from University of Paris-Saclay in 2017, then worked as a postdoctoral researcher at the CMAP (Centre de Mathématiques Appliquées, Ecole Polytechnique). She also did a postdoc at the ICMPE (Institut de Chimie et des Matériaux Paris-Est, CNRS), where she worked on the generation of metal hydrides for hydrogen storage using Machine Learning approaches.

Keywords: Deep Learning, Recurrent Neural Networks, Human-Robot Interaction, emotion recognition, user engagement decrease detection

Jean-Rémy Conti will be focusing on machine learning for spatial data. He will be setting a mathematical framework, in order to establish to what extent the rules obtained by empirical risk minimization can be generalized. It will also serve to study other extensions to the spatial context of popular algorithms, both theoretically and experientially. He is being supervised by Stéphan Clémençon, professor at Télécom Paris and Emilie Chautru, co-director of the geostatistics program and a researcher at the Ecole des Mines de Paris Geostatistics Center.

Keywords: machine learning for spatial data, mathematical, spatial context of popular algorithms, Geostatistics

In 2013, Benoit Fuentes obtained his doctorate at Télécom Paris on audio signal analysis for automatic music transcription and source separation. He then worked for six years at Smart Impulse, a company specialized in power consumption diagnostics to save on energy, where he developed source separation algorithms of electrical signals. Benoit Fuentes was a postdoctoral researcher in the S2A team of the LTCI lab at Télécom Paris from 2020 to 2021. He was developing a generic software tool for deep tensor factorizations and was also working on the underlying probability theory.

Key words: signal processing, source separation, tensor factorization, unsupervised learning, Bayesian inference algorithms.

Go to his website

Robert M. Gower joined Telecom Paris as an Assistant Professor in 2017. He is interested in designing and analyzing new stochastic algorithms for solving big data problems in Machine Learning and scientific computing. A mathematician by training, his academic studies started with a Bachelors and a Master’s degree in applied mathematics at the state University of Campinas (Brazil), where he designed the current state-of-art algorithms for automatically calculating high order derivatives using back-propagation. His PhD in stochastic numeric methods at the University of Edinburgh earned him the 2nd place of the 2017 Leslie Fox prize in numerical analysis. In 2016 he was granted the Fondation Sciences Mathématiques de Paris postdoctoral Laureate fund to continue his work as a postdoc in ENS.

Keywords: Stochastic optimization, randomized numerical linear algebra, convex optimization, machine learning, automatic differentiation.

Visit his website

Key words: survey sampling, transfer learning, nonasymptotic inference, nonparametric statistics.

Halim Hizaoui is an engineer-student at ENSTA Paris on a double-diploma course, ENSA-ENIT (Ecole nationale d’ingénieurs de Tunis). He is passionate about finance and machine learning.

He will be developing a supervised classification method, inspired by the DDaplha classifier (R-package ddalpha). The DDalpha classifier is based on data depth and the alpha-procedure. It offers the benefits of a being non-parametric, fast and robust. The alpha-procedure is an iterative heuristic procedure which results in a discrete space of relevant variables. Function loss optimization takes less computing time thanks to its recursive nature.  The combination of data depth and the alpha-procedure provides promising results, which warrant further development of this method on the heuristics side. The internship will therefore focus on research on the alpha-procedure. Real and synthetic databases will be tested in order to establish the optimal parameters for each configuration and aid decision-making when selecting these parameters. Halim will be supervised by Pavlo Mozharovskyi, associate professor at Télécom Paris.

Orson Jay is a student at Ecole Centrale de Nantes, in Year 1 of the Statistics and Data Science Master’s program. The title of his internship is “Multi-label prediction of social skills in asynchronous video interviews via recurrent neural networks” and is taking place between  01/04/2020 and 28/08/2020. Orson is studying the general skills employers look for in asynchronous video interviews during his internship. He is using a corpus analysis method, studying recent literature on multi-label and label embedding models, and implementing recurrent neural networks based on LSTM or GRU units to solve multi-label problems. Chloé Clavel, teacher-researcher at Télécom Paris and Léo Hemamou, doctoral student at Télécom Paris are supervising the internship.

Key words: automatic natural language processing, deep learning, multi-label models.

Pierre Laforgue is a PhD student in the S²A team of Télécom Paris since October 10th, 2016. A graduate of ENSAE Paris, he also holds the Master 2 "Mathematics, Learning and Human Sciences" from Paris Dauphine University. His thesis, supervised by Florence d'Alché-Buc and Stephan Clémençon, focuses on the unsupervised learning of representations, with the aim of applying it to time series.

Keywords : unsupervised learning, representation learning, time series.

Rémy Leluc’s internship will be on adaptive Monte Carlo methods, in the context of reinforcement learning, examined from a theoretical point of view (convergence and inequalities, theoretical bounds) and from a practical point of view (implementation of new methods and comparisons with state-of-the art methods). He is being supervised by François Portier, lecturer at Télécom Paris and Pascal Bianchi professor at Télécom Paris.

Binh Nguyen is a postdoctoral researcher at S2A team of LTCI from February 2022, working on the application of optimal transport to structured prediction problems in machine learning. He obtained his doctorate degree working with high-dimensional statistics in Proba-Stat team of Département de Mathématiques d’Orsay and INRIA Parietal team. Before that, he studied the Master Data Science at Paris-Saclay University.

Keywords: optimal transport, structured prediction, high-dimensional statistics, sparisity

Personal website: https://tbng.github.io/

Nathan Noiry is a postdoctoral researcher in the LTCI S2A Research Team (Signal, Statistics and Learning). He is working on survey sampling, transfer learning and broadly speaking on machine learning problematics. After studying at Ecole Normale Supérieure de Lyon, he completed a PhD in probability theory, during which he worked on random matrices and random graphs.

Key words: machine learning, survey sampling, transfer learning, random matrices, random graphs

Go to his website

Brice Parilusyan is passionate about robotics. He joined his school’s robotics association, twice taking part in the French robotics championship and built InMoov, a life-size humanoid robot. His internship will be on the study of social signals in a group. He is studying the “emerging state” in a group, such as cohesiveness and leadership and how they can be quantified by a virtual agent. He is being supervised by Giovanna Varni, lecturer at Télécom Paris.

François Portier is a lecturer in the LTCI S2A team at Télécom Paris. His PhD at Université de Rennes 1 was on parsimonious predictive models. He was then an FNRS postdoc at Université catholique de Louvain, where he studied certain survival analysis models. His current research work is on sequential Monte-Carlo methods, machine learning for censored and dependent data and parsimonious predictive models.

Keywords: adaptive sampling, MCMC, bootstrap, Markov chain, dimensionality reduction, survival analysis.

Visit his website

Oskar Rynkiewicz is an IASD Master's student at Paris Dauphine-PSL university. He takes an interest in optimization and machine learning. At Télécom Paris, he joined the S²A team from 30/03/2020 to 30/09/2020 to undertake his research internship under the supervision of Olivier Fercoq. He seeks to prove a lower complexity bound of primal-dual algorithms for convex affinely constrained optimization problems under metric sub-regularity. Once obtained, the lower bound will verify the optimality of the currently used methods with respect to metric sub-regularity. He holds an engineering degree from IMT Atlantique.

keywords : convex optimization, lower bound, rate of convergence, metric sub-regularity, machine learning

Anne Sabourin is assistant professor in the S2A research team at Télécom Paris since 2013. She received her PhD degree in 2013 on multivariate extreme values and Bayesian inference at Lyon 1 University, under the supervision of Anne-Laure Fougères and Philippe Naveau. Her research interests concern multivariate extreme value theory, dependence between rare events, dimension reduction in extreme regions, with various applications ranging from environmental risk to machine learning applications such as anomaly detection.

Keywords: multivariate extreme value theory, rare events, dimension reduction, anomaly detection.

Visit her website

Othmane Sebbouh is a Data Science Master's student at ENSAE & Ecole polytechnique. The topic of his internship at the DSAIDIS chair is "Towards closing the gap between the theory and practice of stochastic variance reduced methods". SVRG is an inner-outer loop based method, where in the outer loop a reference full gradient is evaluated, after which m steps of an inner loop are executed where the reference gradient is used to build a variance reduced estimate of the current gradient. The simplicity of the SVRG method and its analysis has lead to multiple extensions and variants for even non-convex optimization. Yet there is a significant gap between what parameter settings the analysis suggests and what is known to work well in practice, which is why Othman's work will be to take several steps towards closing this gap.

Umut Şimşekli is an assistant professor at Télécom Paris. He received his PhD degree in 2015 on inference methods for large-scale matrix and tensor factorization models in the Department of Computer Engineering at Bogaziçi University, İstanbul, Turkey. His research interests are in scalable Bayesian machine learning, audio and music processing, and recommendation systems.

Keywords: matrix and tensor factorizations, Markov Chain Monte Carlo, audio and music processing.

Visit his website

Yousef Taheri Sojasi is passionate about machine learning and natural language processing. His internship, “Weak signals and text data” is supervised by Stephan Clémençon, professor at Télécom Paris and Matthieu Labeau, associate professor at Télécom Paris. The internship started on 30/03/2020 and will end on 31/08/2020. Its aim is the development of representation methods that  make it easier to detect weak signals in text data. Weak signal detection is a major challenge when it comes to applications.  The method takes its inspiration from methods and criteria based on extreme value theory, which extend the scope of supervised and unsupervised learning techniques.

Key words: automatic natural language processing, word representation, extreme value theory, supervised learning, unsupervised learning

Giovanna Varni joined Télécom Paris as an Assistant Professor in 2017. Her research activities focus on the field of socio-affective computing. Previously, Giovanna Varni was a postdoctoral researcher at the University of Genoa (Italy) on the InfoMus Lab team and then in the INTERACTION team (IMI2S group) at the Pierre et Marie Curie University in Paris 6. Her work focuses on the analysis of non-verbal multimodal signals in human-human interaction and human-machine interaction. Since 2006, she has been involved in several European projects (FP7, EU-ICT, STREP and FET).

Keywords: human-human and man-machine interaction, socio-affective signal processing, interpersonal synchrony, expressive gesture.

Visit her website

Robin Vogel completed his engineering degree from ENSAE Paris as well as the M2 Data Science from Université Paris-Saclay in 2016. He was a PhD student at Télécom Paris from 2017 to 2020, under the supervision of Stephan Clémençon, Aurélien Bellet (Inria) and Anne Sabourin. He was employed by the leading biometrics and security solutions company IDEMIA, through a CIFRE agreement. He was interested in statistical learning theory and machine learning for identification and access control.

Keywords: similarity learning, ranking, biometrics, statistical learning theory, deep learning.

Visit his website