Applications are invited for 2 Post-doc positions available in the period 2020- 2023.
The broad work area is “AI/Machine Learning for Graphs and Text”.More specifically topics are:
• deep learning for text streams – the case of # prediction in twitter.
• text generation algorithms for document understanding
• deep learning for graph & graph node embeddings.
The candidates must have at least two of the following skill sets:
• Sound mathematical background (Proba/Stat, Linear Algebra,)
• Experience in Text Retrieval & Mining (recommendation algorithms, text categorization
opinion/sentiment mining, text generation)
• Deep Learning skills (with recent architectures)
• Very good programming skills (including Python)
- a recent Ph.D. degree in either: Computer Science, Applied Mathematics, Physics.
- analytical skills and creative thinking with a hard working attitude
- a sound publication record with visible impact
Funding: Competitive funding (salary, travel budget, budget for interns) for up to 36 months.
Interested graduate students should send by email to Mrs. Jessica GAMEIRO (firstname.lastname@example.org)
- a cover letter including presentation of their academic record, motivation of the candidate.
- a full CV with detailed grading information for the acquired degrees.
The postdoctoral researchers will be based in the Informatics Laboratory (LIX) of EcolePolytechnique in the broader area of Paris. Ecole Polytechnique is the premier engineering University of France (highly ranked internationally according to the latest University rankings). Famous scientists (including Nobel prize and Field medal recipients) and industrial leaders are alumni of the school, offering an exceptional environment for research in the centre of the fast growing excellence pole of Saclay few km south of Paris.Ecole Polytechnique is as well the main founding member of the new INSTITUT POLYTECHNIQUE DE PARIS federating some of the best engineering schools of France. The DaSciM group has already a significant impact in local and international research and industrial activities. Additionally, it offers ample computing resources and facilities in the University campus. See further details at DASCIM group: http://www.lix.polytechnique.fr/dascim/
Several teaching positions at École Polytechnique are currently open. The deadline for applications is March 23, 2020.
All the details concerning the positions and the procedure to apply are accessible from the DIX website.
Pour le prochain séminaire de l’équipe Cosynus, nous aurons le plaisir d’accueillir Marie Kerjean pour son exposé intitulé Typing differentiable programming.
Résumé: Differentiable programming is a recent research area: its objective is to express differentiation as a modular algorithmic transformation on rich programming languages. It is in particular motivated by the various applications of automatic differentiation in machine learning or formal calculus. In this talk I will present joint work with Pierre-Marie Pédrot, focusing on the typing system used to express differentiation.
We will first review a few examples of differentiable languages recently exhibited in the literature. This allows to identity the linear Dialectica transformation as a reverse automated differentiation transformation on a higher-order lambda-calculus with positive types. Building on the intuitions provided by Dialectica and distribution theory, we construct a lambda-calculus with an internal differentiation operator. This calculus is typed by a type system inspired by Differential Linear Logic. Noticeably, we are able to express backward automatic differentiation as a call-by-name strategy and forward automatic differentiation as a call-by-value strategy.
Mathieu Carrière, currently PostDoc at Columbia University, will give a talk next Monday 24th at 2:30pm in Gilles Kahn, on Statistics and Machine Learning in Topological Data Analysis with Applications to Biology.
Abstract: Topological Data Analysis (TDA) is an area of data science which aims at characterizing data sets with their topological features in various dimensions. Examples of such features include the connected components, the loops or the cavities that are present in the data, and which are encoded in the two main descriptors of TDA, the so-called persistence diagram and Mapper. Even though these descriptors have proved useful in many applications, it is not straightforward to include them in automated processes, which are common in statistics and machine learning, mainly because the space of these descriptors lacks a lot of required properties, such as a well-defined addition or barycenter. In this talk, I will recall the basics of TDA and review the current solutions that have been proposed in the past few years to merge TDA descriptors (persistence diagram and Mapper) with statistics and Machine Learning. Then, I will introduce some questions that remain open in this topic, and that are active fields of research as of today, such that the question of persistence diagram differentiability for deep learning, or the statistical analysis of the Mapper in the multivariate case. In the process, I will also illustrate these problems by providing applications on biological data, such as immunofluorescence images for breast cancer pathology and single-cell RNA sequencing for understanding the spinal cord cellular diversity.
There will be a seminar on Thursday 6 February at 14:30 at LIX, room Grace Hopper with a talk given by Nguyễn Kim Thắng (IBISC).
Title: Online Non-monotone Submodular Maximization: Approximation and Regret Guarantees
Abstract: Submodular and diminishing-returns (DR) submodular optimization are important optimization problems with many real-world applications in machine learning, economics and communication systems. Moreover, DR-submodular optimization captures a subclass of non-convex optimization that provides both practical and theoretical guarantees.
In this talk, we present algorithms with performance guarantees for the fundamental problems of maximizing non-monotone submodular and DR-submodular functions over convex sets in the online environment. In several settings, these guarantees match the currently best-known ones in the offline environment.
Dr. Johannes Lutzeyer, new postdoctoral researcher in DaSciM, will give a presentation at 14:30 entitled: Extending the Davis-Kahan theorem for the comparison of embedding spaces spanned by eigenvectors
Abstract: In this talk I will introduce the Davis-Kahan theorem, which is commonly used to upper bound the distance of two embedding spaces. It often forms an essential part of consistency proofs of algorithms based on spectral graph embeddings and variants of principal component analysis (PCA). However, the Davis-Kahan theorem has several weaknesses, which is why work on it continues. In this talk I present an extended version of the Davis-Kahan theorem which addresses these weaknesses and give a few proof of concept examples such as covariance embedding spaces in the context of the PCA algorithm and graph shift operator embedding spaces in the context of the spectral clustering algorithm. The work presented is joint work with Andrew Walden.
La prochaine séance du séminaire Combi du Plateau de Saclay aura lieu ce mercredi à 10h30 dans la salle Philippe Flajolet du LIX. Nathan Noiry (Modal’X, Université Paris Nanterre) nous parlera de «L’algorithme de parcours en profondeur dans un modèle de configuration».
Le programme du séminaire est disponible ici : https://galac.lri.fr/pages/combi-seminar.html
Résumé: Dans cet exposé, issu d’un travail en collaboration avec Nathanaël Enriquez, Gabriel Faraud et Laurent Ménard, nous nous intéresserons à des graphes aléatoires dont la suite des degrés est fixée. Nous verrons que ce modèle présente une transition de phase concernant l’existence d’une composante connexe de taille macroscopique. Dans le régime surcritique, nous nous intéresserons à l’algorithme de parcours en profondeur sur cette composante géante, qui en construit un arbre couvrant. Notre résultat principal établit la convergence du processus de contour renormalisé associé à cet arbre, vers un profil explicite. Une conséquence de ce résultat est l’existence de chemins simples macroscopiques dans le graphe. Nous aborderons ensuite quelques éléments de preuve et verrons qu’au cours de l’exploration, l’évolution de la mesure empirique des degrés au sein du graphe induit par les sommets non-explorés admet une limite fluide. Celle-ci est décrite par un système infini d’équations différentielles qui, de manière inattendue, admet une solution unique et explicite en fonction de la série génératrice de la loi initiale des degrés.
Marie-Paule Cani, professeure à l’École polytechnique et membre de l’équipe GeoViC, est l’une des 18 nouveaux membres élus à l’Académie des sciences lors de la session d’élection débutée en 2019. L’Académie des sciences est une institution indépendante, placée sous la protection du président de la République, qui rassemble des personnalités scientifiques, choisies parmi les plus éminents spécialistes français et étrangers. Elle participe à la promotion et au développement des sciences et assure un rôle d’expertise et de conseil auprès des autorités gouvernementales.