I am an associate professor of Artificial Intelligence in the Department of Computer Science at Aix-Marseille University and the Machine Learning group QARMA at LIS Lab.
Contact Information
- hachem (dot) kadri (at) univ-amu (dot) fr
- (+33) (0)4.91.05.46.52
- Parc Scientifique et Technologique de Luminy
163, avenue de Luminy – Case 901
F-13288 Marseille Cedex 9, France
Research Interests
My research interests in machine learning include kernel methods, functional data analysis, statistical learning theory, deep learning, and more recently quantum machine learning.
News:
- Nominated member of the National Universities Council (CNU, 2024-2027) in section 27 (Computer Science).
- Elected in 2023 as Head of Data Science Department at the Computer Science Lab LIS of Aix-Marseille University.
- Two conference papers published: « Implicit Regularization in Deep Tucker Factorization: Low-Rankness via Structured Sparsity » at AISTATS-2024, and « Deep Learning with Kernels through RKHM and the Perron-Frobenius Operator » at NeurIPS-2023
- A paper on « Learning in RKHM: a C*-algebraic twist for kernel machines » got accepted at AISTATS-2023.
- We are organizing the first workshop on Quantum Machine Learning at ECML-PKDD 2022.
- A paper on « Quantum Perceptron Revisited: Computational-Statistical Tradeoffs » got accepted at UAI-2022.
- A paper on « Implicit Regularization with Polynomial Growth in Deep Tensor Factorization » got accepted at ICML-2022.
- Vice-president of SSFAM – Société Savante Francophone d’Apprentissage Machine.
- Two papers on « Entangled Kernels – Beyond Separability » and « Toolbox for Multimodal Learn (scikit-multimodallearn) » got accepted at Journal of Machine Learning Research (JMLR).
- QuantML — Quantum Machine Learning: Foundations and Algorithms — ANR Starting Grant (JCJC) project accepted for 2019 – 2023 (PI).
- OPERA — Optimization and Compression of Deep Learning Models — Franco-Tunisian PHC Utique project accepted for 2019 – 2022 (co-PI).
- On June 27th, 2019 at Aix-Marseille University, I defended my HDR thesis on « Learning with Operator-valued Kernels: Foundations and Algorithms ».
The HDR (Habilitation à diriger des recherches) is the highest degree in the French academic system.
HDR Committee: Florence d’Alché-Buc, Stéphane Canu, Frédéric Ferraty, Liva Ralaivola, John-Shawe Taylor, Bruno Torrésani, Jean-Philippe Vert.
From September 2012 to June 2019, I was an Assistant Professor in the Department of Computer Science at Aix-Marseille University. I was also a member of the Qarma (Machine Learning and Multimedia) team which is situated within the Fundamental Computer Science laboratory (Laboratoire d’Informatique Fondamentale de Marseille, a.k.a. LIF). My research interests lie principally in the fields of machine learning, statistics, and signal processing, more precisely in: kernel methods, functional data analysis, and speech and audio processing. Although I have been working and continue to work on research issues specific to each of these fields, many of my interests lie in the intersection of them, and cover both methodological as well as application aspects.
From December 2009 to August 2012, I was a postdoctoral researcher in the SequeL (Sequential Learning) team at INRIA Lille – Nord Europe in France working with Philippe Preux on regularization for functional regression models: theory, sequential algorithms and applications.
From November 2008 to December 2009, I was a postdoctoral fellow at the LAGIS (Laboratory of Control, Computer Science & Signal) in Lille, France. the research project that I worked on, under the supervision of Emmanuel Duflos and the collaboration of Manuel Davy, involved functional data analysis: classification and regression.
In 2008, I received my Ph.D. in Electrical Engineering from the National Engineering School of Tunis (ENIT), working with Systems and Signal Processing Lab (LSTS) and the High School of Technology and Computer Science (ESTI). My advisor was Pr. Noureddine Ellouze. My research has explored a variety of topics in statistical audio signal processing, including speaker segmentation (speaker change detection and clustering), audio classification (impulsive sounds classification), and the application of these tasks to audio content indexing and retrieval.