Hi, bonjour, yeia

Welcome to my personal web-page. I am Argyris Kalogeratos (Αργύρης Καλογεράτος), I come from Patras, Greece. My background is on Computer Science and I have a PhD on Machine Learning (2013).

Click for short bio

Argyris Kalogeratos has BSc and MSc degrees in Computer Science from the University of Ioannina, Greece. He holds a PhD in Machine Learning and Information Technologies (2013, supervisor: Prof. Aristidis Likas), from the same Institution. He has been a research scientist in the Centre of Applied Mathematics, ENS Cachan, France (2013-2018), and since 2018 he is a staff researcher in the Centre Borelli, ENS Paris-Saclay, Université Paris-Saclay, France. His interests include Machine Learning, Statistics, Network Science, Decision & Control, and Computational Social Systems. He has been involved in several synergistic projects between Academia and Industry, for problems related to multimedia mining, social network analysis and mining, computational epidemiology, marketing, transportation, communication systems, and bioinformatics. Argyris likes multidisciplinary research and the fusion of ideas, concepts, viewpoints, originating from different domains.

Currently

Position: I am a permanent researcher at the Centre Giovanni Borelli (ex Centre de Mathématiques et de Leurs Applications – CMLA), Ecole Normale Supérieure Paris-Saclay (ENS Paris-Saclay), University Paris-Saclay, France, and member of the Machine Learning and Massive Data Analysis (MLMDA) research group.

Research focus: Coordinator of the ML on Graphs research theme.

Research initiatives: I participate to the research activities of the ML Industrial Big Data Chair. Among our industrial partners, there are SNCF, CEA, Michelin.

Past funding: My research position at CMLA has been funded in the past by the MORANE project (in collaboration with SNCF railway company) and the SODATECH project  (funded by the French state) for which I was the research coordinator and responsible for CMLA’s part. See more details at the “Projects” section.

Centre Borelli in a sentence

CB is a unique and dynamic environment where the academic research is developed while being challenged in real applications.

 

 

Announcements

EDF Data Challenge

The EDF Data Challenge is still open for contributions on the problem of forecasting electricity load (consumption). Feel free to diffuse and contribute! This is a data challenge in which our student Eloi Campagne leads the organization team. A paper giving more technical details can be found here.

Research topics for ARIA students 2024

Here are some subjects that could be investigated by project as part of the ARIA internship 2023-2024.

— Subject 1 — Online stream k-diversification

— Subject 2 — Outlier detection for graph signals

— Subject 3 — Graph representation learning

— Subject 4 — Graph-theoretic pooling for Graph Neural Networks

— Subject 5 — Electricity load forecasting
Info

Material: [1][2][3].

In addition to the above, and depending on the background and the interests of a student, there can be other possibilities to define a project after some discussion. Here are some topics that could be interesting:

  • Graph theoretical Machine Learning
  • Data clustering
  • Statistical testing in high dimensions
  • Interpretable/Explainable Machine Learning
  • Social network analysis and diffusion of information
  • Problems related to sequential decision-making
  • Problems related to epidemic spreading and control
  • Problems about urbanism and mobility (e.g. transportation, public safety, etc)

 

News

  • EDF Challenge: The EDF Data Challenge (our student Eloi Campagne leads the organization team) is still open for contributions on the problem of forecasting electricity load (consumption). Feel free to diffuse and contribute!
  • SETN2024: A new paper with the title LIPO+: Frugal Global Optimization for Lipschitz Functions, co-authored with G. Serré (PhD student, CB), P. Beja-Battais (PhD student, CB), S. Chirrane, and N. Vayatis, will appear in EETN Conference on Artificial Intelligence. [pdf]
  • W@ECML2024: A new paper with the title Leveraging Graph Neural Networks to Forecast Electricity Consumption, co-authored with Eloi Campagne (PhD student, EDF&CB), Yvenn Amara-Ouali, and Yannig Goude, will appear at the European Conference on Machine Learning – Workshop ML4SPS (with proceedings), 2024. [pdf]
  • JdS 2024: Eloi Campagne (PhD student, EDF&CB) presented his work on the Forecasting Net Load in France – EDF Challenge 2024, at the 55th Journées de Statistique (JdS 2024, SFdS), in May, Bordeaux, France. Joint work with Yvenn Amara-Ouali and Yannig Goude from EDF. [pdf]
  • New interns:
    • Louise Allain (M2 MVA, Apr–Oct) on non-parametric graph inference.
    • Silviu-Adrian Predoi (M1 ENS P.S. Physics) on node embeddings (Apr-Sep).
    • Tokinirina Hansen and Yorick Libiot (L3 ENS P.S. Maths) on random walks over graphs.
  • Preprint: A new priprint with the title Collaborative non-parametric two-sample testing, co-authored with A.de la Concha, and N. Vayatis, 2024. [pdf]
  • Preprint: A new priprint with the title Stein Boltzmann Sampling: A Variational Approach for Global Optimization, co-authored with G. Serré, and N. Vayatis, 2024. [pdf]
  • AISTATS 2024: The paper Online non-parametric likelihοod-ratio estimation by Pearson-divergence functional minimization. with A. de la Concha and N. Vayatis, will appear in the 27th International Conference on Artificial Intelligence and Statistics (AISTATS), 2024 [pdf].
  • Preprint: Significant update to the work Collaborative likelihood-ratio estimation over graphs, co-authored with A. de la Concha, and N. Vayatis, can be found at [pdf].
  • Preprint: The new preprint UniForCE: The unimodality forest method for clustering and estimation of the number of clusters, co-authored with G. Vardakas and A. Likas, can be found at [pdf][code].
  • Preprint: The new preprint Online non-parametric likelihood-ratio estimation by Pearson-divergence functional minimization, co-authored with A. de la Concha and N. Vayatis, can be found at [pdf].
  • New interns:
    • Jeanne Belly (M1 ENS P.S. ARIA) on social network study of propagation phenomena.
    • Teo Tessier (M1 ENS P.S. ARIA) on graph generation methods.
  • ΤCSS: The article DeGroot-based opinion formation under a global steering mechanism, co-authored with I. Conjeaud and P. Lorenz-Spreen, has been accepted to the journal IEEE Transactions on Computational Social Systems (TCSS, impact factor: 4.75) [pdf].
  • RTE-SNCF: Launch of a new ambitious project with RTE and SNCF, also in collaboration with Eurobios. Machine learning, operational research, and network science, toward modern scalable predictive maintenance.
  • CNA2023: An extended abstract under the title Opinion formation under global steering with application to social network data analysis will be presented at the Complex Networks 2023 conference [pdf][slides]. This refers to the following long full paper that is co-authored with Ivan Conjeaud and Philipp Lorenz-Spreen.
  • ICTAI2023: The paper A framework for paired-sample hypothesis testing for high-dimensional data, co-authored with I. Bargiotas and N. Vayatis, will appear in the IEEE International Conference on Tools with Artificial Intelligence 2023 [pdf].
  • New PhD students:
    • Gaëtan Serré, co-supervision with Nicolas Vayatis
    • Gwendal Debaussart
    • Gaspard Abel, co-supervision with Julien Randon-Furling and Jean-Pierre Nadal.
  • ARIA topics: Some topics are now available for the new students of the ARIA program.
  • Invited Talk: At the 10th International Workshop on Applied Probability (IWAP) 2023 at Thessaloniki, I talked about the Warm-starting selection process and  its multi-round extention.
  • New interns:
    • Perceval Beja-Battais (M2 MVA, Apr – Oct) will work on learning theory.
    • Gaëtan Serré (M2 MVA, Apr – Oct) will work on global optimization.
    • Gwendal Debaussart (M2 MVA, Apr – Oct) will work on statistical machine learning.
    • Eloi Campagne (M2 MVA, Apr – Oct) will work on time-series analysis and prediction (colab. with EDF).
    • Gaspard Abel (M2 ENS PS Physics) on information diffusion, co-supervision with Jean-Pierre Nadal.
    • Mohamed Aymen Bouyahia (M1 ENSTA Paris, May – August) will work on interpretable ML.
    • Abdeljalil Zoubir (UM6P PhD visiting as an intern from Morocco, Apr – July) on graph neural networks with application to chemical compounds.
    • Manon Gouttefangeas and David Pieroucci (L3 ENS Maths, Apr – Jun) on paired sample hypothesis testing.
    • Quentin Boiret and Amer Essakine (L3 ENS Maths, Apr – Jun) on graph neural networks for time series.

See all news...