Publications
Published
Distributionally Robust Optimization with Bias and Variance Reduction.
Ronak Mehta, Vincent Roulet, Krishna Pillutla, Zaid Harchaoui.
International Conference on Learning Representation (ICLR), 2024.
paper
On the Interplay Between Stepsize Tuning and Progressive Sharpening.
Vincent Roulet, Atish Agarwala, Fabian Pedregosa.
NeurIPS OPT 2023: Optimization for Machine Learning, 2023.
paper
Modified Gauss-Newton Algorithms under Noise.
Krishna Pillutla, Vincent Roulet, Sham Kakade, Zaid Harchaoui.
2023 IEEE Statistical Signal Processing Workshop (SSP), 2023.
paper
Stochastic Optimization for Spectral Risk Measures.
Ronak Mehta, Vincent Roulet, Krishna Pillutla, Lang Liu, Zaid Harchaoui.
Proceedings of the 26th International Conference on Artificial Intelligence and Statistics, 2023.
paper code
Target Propagation via Regularized Inversion for Recurrent Neural Networks.
Vincent Roulet, Zaid Harchaoui.
Transactions in Machine Learning Research, 2023.
paper code one slide summary
Revisiting Convolutional Neural Networks from the Viewpoint of Kernel-Based Methods.
Corinne Jones, Vincent Roulet, Zaid Harchaoui.
Journal of Computational and Graphical Statistics, 2022.
paper code
A Representation-Focused Training Algorithm for Deep Networks.
Vincent Roulet, Corinne Jones, Zaid Harchaoui.
2022 IEEE Data Science and Learning Workshop (DSLW), 2022.
paper long version one slide summary slides
Differentiable Programming à la Moreau.
Vincent Roulet, Zaid Harchaoui.
Proceedings of the International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2022.
paper one slide summary poster slides
Discriminative Clustering with Representation Learning with any Ratio of Labeled to Unlabeled Data.
Corinne Jones, Vincent Roulet, Zaid Harchaoui.
Statistics and Computing, 2022.
paper code
On the Smoothing of Deep Networks.
Vincent Roulet, Zaid Harchaoui.
In Proceedings of the 55th Annual Conference on Information Sciences and Systems (CISS), 2021.
paper
On the Convergence to Stationary Points of the Iterative Linear Exponential Quadratic Gaussian Algorithm.
Vincent Roulet, Maryam Fazel, Siddhartha Srinivasa, Zaid Harchaoui.
In Proceedings of the 2020 American Control Conference (ACC), 2020.
paper code
End-to-End Learning, with or without Labels.
Corinne Jones, Vincent Roulet, Zaid Harchaoui.
Short version in Proceedings of Joint Statistical Meeting (JSM), 2020.
Winner of the 2020 ASA Computing/Graphics Student Paper Award.
paper code poster
Sharpness, Restart and Acceleration.
Vincent Roulet, Alexandre d’Aspremont.
(Journal version) SIAM Journal of Optimization, 2020.
paper
An Elementary Approach to Convergence Guarantees of Optimization Algorithms for Deep Networks.
Vincent Roulet, Zaid Harchaoui.
Proceedings of the 57th Annual Allerton Conference on Communication, Control, and Computing (Allerton), 2019.
paper
Iterative Linearized Control: Stable Algorithms and Complexity Guarantees.
Vincent Roulet, Siddharta Srinivasa, Dmitry Drusvyatskiy, Zaid Harchaoui.
Proceedings of the 36th International Conference on Machine Learning (ICML), 2019.
paper code poster
Computational Complexity versus Statistical Performance on Sparse Recovery Problems.
Vincent Roulet, Nicolas Boumal, Alexandre d’Aspremont.
Information and Inference: a Journal of the IMA, 2019.
paper
A Smoother Way to Train Structured Prediction Models.
Krishna Pillutla, Vincent Roulet, Sham Kakade, Zaid Harchaoui.
Advances in Neural Information Processing Systems 31 (NeurIPS), 2018.
paper code poster blog post
Sharpness, Restart and Acceleration.
Vincent Roulet, Alexandre d’Aspremont.
(Conference version) Advances in Neural Information Processing Systems 30 (NeurIPS), 2017.
paper
Integration Methods and Accelerated Optimization Algorithms.
Damien Scieur, Vincent Roulet, Francis Bach, Alexandre d’Aspremont.
Advances in Neural Information Processing Systems 30 (NeurIPS), 2017.
paper
Working papers
The Elements of Differentiable Programming.
Mathieu Blondel, Vincent Roulet.
book
Stepping on the Edge: Curvature Aware Learning Rate Tuners.
Vincent Roulet, Atish Agarwala, Jean-Bastien Grill, Gregor Swirszcz, Mathieu Blondel, Fabian Pedregosa.
paper
Dual Gauss-Newton Directions for Deep Learning.
Vincent Roulet, Mathieu Blondel
paper
Complexity Bounds of Iterative Linear Quadratic Optimization Algorithms for Discrete Time Nonlinear Control.
Vincent Roulet, Siddharta Srinivasa, Maryam Fazel, Zaid Harchaoui.
paper slides poster
Iterative Linear Quadratic Optimization for Nonlinear Control: Differentiable Programming Algorithmic Templates.
Vincent Roulet, Siddharta Srinivasa, Maryam Fazel, Zaid Harchaoui.
Companion report to Complexity Bounds of Iterative Linear Quadratic Optimization Algorithms for Discrete Time Nonlinear Control.
paper code notebook tutorial
PhD Thesis
On the Geometry of Optimization Problems and their Structure.
Vincent Roulet.
PhD Thesis.
manuscript slides