Email: joseph.salmon@umontpellier.fr
Address: IMAG, c.c. 051 Université de Montpellier Place Eugène Bataillon 34095 Montpellier Cedex 5
(office 415, building 9)
Phone: +33 4 67 14 35 19
Recent Publications
Concomitant Lasso with Repetitions (CLaR): beyond averaging multiple realizations of heteroscedastic noise
Q. Bertrand, M. Massias, A. Gramfort and J. Salmon (2019).
[Code]
[BibTeX]
@techreport{Bertrand_Massias_Gramfort_Salmon19,
author = {Q. Bertrand, M. Massias, A. Gramfort and J. Salmon},
comment = {
[Code] },
pdf = {https://hal.archives-ouvertes.fr/hal-02010014/document},
title = {Concomitant Lasso with Repetitions (CLaR): beyond averaging multiple realizations of heteroscedastic noise},
year = {2019}
}
Screening Rules for Lasso with Non-Convex Sparse Regularizers
A. Rakotomamonjy, G. Gasso and J. Salmon (2019).
[BibTeX]
@techreport{Rakotomamonjy_Gasso_Salmon19,
author = {A. Rakotomamonjy, G. Gasso and J. Salmon},
pdf = {https://hal.archives-ouvertes.fr/hal-02015085v1},
title = {Screening Rules for Lasso with Non-Convex Sparse Regularizers},
year = {2019}
}
Optimal mini-batch and step sizes for SAGA
N. Gazagnadou, R. Gower and J. Salmon (2019).
[BibTeX]
@techreport{Gazagnadou_Gower_Salmon19,
author = {N. Gazagnadou, R. Gower and J. Salmon},
pdf = {https://arxiv.org/pdf/1902.00071.pdf},
title = {Optimal mini-batch and step sizes for SAGA},
year = {2019}
}
A hierarchical Bayesian perspective on majorization-minimization for non-convex sparse regression: application to M/EEG source imaging
Y. Bekhti, F. Lucka, J. Salmon and A. Gramfort (2018).
Inverse Problems
34: (085010).
[BibTeX]
@article{Bekhti_Lucka_Salmon_Gramfort18,
author = {Y. Bekhti, F. Lucka, J. Salmon and A. Gramfort},
journal = {Inverse Problems},
number = {8},
pages = {085010},
pdf = {https://arxiv.org/abs/1710.08747},
title = {A hierarchical Bayesian perspective on majorization-minimization for non-convex sparse regression: application to M/EEG source imaging},
volume = {34},
year = {2018}
}
Statistical Inference with Ensemble of Clustered Desparsified Lasso
J.-A. Chevalier, J. Salmon and B. Thirion (2018).
MICCAI
[BibTeX]
@inproceedings{Chevalier_Salmon_Thirion18,
author = {J.-A. Chevalier, J. Salmon and B. Thirion},
booktitle = {MICCAI},
pdf = {https://arxiv.org/pdf/1806.05829.pdf},
title = {Statistical Inference with Ensemble of Clustered Desparsified Lasso},
year = {2018}
}
On Lasso refitting strategies
E. Chzhen, M. Hebiri and J. Salmon (2018).
Bernouilli (to appear)
.
[BibTeX]
@article{Chzhen_Hebiri_Salmon17,
author = {E. Chzhen, M. Hebiri and J. Salmon},
journal = {Bernouilli (to appear)},
pdf = {https://arxiv.org/pdf/1707.05232.pdf},
title = {On Lasso refitting strategies},
year = {2018}
}
Dual Extrapolation for Faster Lasso Solvers
M. Massias, A. Gramfort and J. Salmon (2018).
ICML
[Python Code] |
[BibTeX]
@inproceedings{Massias_Gramfort_Salmon18,
author = {M. Massias, A. Gramfort and J. Salmon},
booktitle = {ICML},
comment = {
[Python Code] |},
pdf = {https://arxiv.org/abs/1802.07481},
title = {{Dual Extrapolation for Faster Lasso Solvers}},
year = {2018}
}
Contact
News
- Workshop in Montpellier (May 2-3 2019): Graph signals : learning and optimization perspectives
- January 2019: New preprint out, joint work with Q. Bertrand, M. Massias and A. Gramfort. "Concomitant Lasso with Repetitions (CLaR): beyond averaging multiple realizations of heteroscedastic noise" arxiv.
- January 2019: New preprint out, joint work with N. Gazagnadou, R. Gower and J. Salmon. "Optimal mini-batch and step sizes for SAGA" arxiv.
- January 2019: New preprint out, joint work with G. Gasso and A. Rakotomamonjy. " Screening Rules for Lasso with Non-Convex Sparse Regularizers " arxiv
- October 2018: New preprint out, joint work with E. Ndiaye, T. Le, O. Fercoq, I. Takeuchi : on links between optimization and learning "Safe Grid Search with Optimal Complexity" pdf.
- July 2018: "Celer: a Fast Solver for the Lasso with Dual Extrapolation", accepted to ICML2018; code and companion paper called (pdf, slides). Joint work with Mathurin Massias and Alexandre Gramfort
- New paper "Statistical Inference with Ensemble of Clustered Desparsified Lasso" with Jérôme-Alexis Chevalier and Bertrand Thirion accepted at Miccai 2018 pdf
- classgrade: an open source project for peer grading, mostly developped by C. Marini. In the process of being merged into: Peeramid.
Short Bio
I am currently a full professor at Université de Montpellier and an associate member at INRIA Parietal Team. For the spring and summer quarters 2018, I was a visiting assistant professor at UW, Statistics departement. From 2012 to 2018 I was an assistant professor at Telecom ParisTech. Back in 2011 and 2012, I was a post-doctoral Associate at Duke university working with Rebecca Willett.
In 2010, I finished my PHD in statistics and image processing under the supervision of Dominique Picard and Erwan Le Pennec at the Laboratoire de Probabilités et de Modélisation Aléatoire, now LPSM, in Université Paris Diderot.
Software
- matlab toolboxes for statistics and image processing (this is legacy)
- Celer: a fast Lasso solver, code of the associated ICML paper "Dual Extrapolation for Faster Lasso Solvers" (pdf, slides)
More on my Github Page
Team
PhD Students
- Mathurin Massias (co-advised with Alexandre Gramfort): 2016-2019?
- Jérôme-Alexis Chevalier co-advised by Bertrand Thirion 2017-2020?
- Evgenii Chzhen co-advised by Mohamed Hebiri 2016-2019?
- Quentin Bertrand co-advised by Alexandre Gramfort) 2018-2021?
- Nidham Gazagnadou co-advised by Robert Gower) 2018-2021?
Alumni
- Eugene Ndiaye [PhD, 2015-2018], co-advised by Olivier Fercoq
- Jean Lafond [PhD, 2013-2016] co-advised by Éric Moulines, (now at Cubist Systematic)
- Igor Colin [PhD, 2013-2016] co-advised by Stéphan Clémençon and founded by Streamwide (now at Huawei)
- Jair Montoya [Post Doc, 2016-2017], co-advised by Olivier Fercoq
- Thierry Guillemot (now at Rythm Inc.) [Engineer, co-advised by A. Gramfort, 2016]
Open positions in my lab
This list is fuzzy so please contact me directly for potential opportunities (post-doc, phd thesis, internship)