Posts

My paper on Regularized Newton got accepted for publication at SIAM Journal on Optimization (SIOPT). The main result of this work is to show that one can globalize Newton’s method by using regularization proportional to the square root of the gradient norm. The corresponding method achieves global acceleration over gradient descent and it converges with the $O(1/k^2)$ rate of cubic Newton.

I’m taking part in the 2022 Workshop on Federated Learning and Analytics on 9 and 10 November. I am giving a talk about our work on Asynchronous SGD in the mini-workshop on Federated Systems at Scale on 9 November.

I’m giving a talk at Séminaire Parisien d’Optimisation on 10 October. I will be presenting my work on second-order optimization, including the Super-Universal Newton paper.

This week I’m participating in the Learning and Optimization in Luminy (LOL2022) workshop. Yesterday I gave a short presentation of our Super-Universal Newton paper.

We recently released two new papers that are available on arxiv as preprints. The first work considers the global convergence of regularized Newton method (short summary in this twitter thread) and the second work is about adaptive stepsizes for stochastic gradient descent.

Between 22 June and 28 June, I visited Sebastian Stich, who works at CISPA Helmholtz Center for Information Security near Saarbrücken in Germany. We discusses new ways to solve nonconvex problems, and we already have some interesting results. Sebastian is also looking for PhD students, so if you or someone you know is looking for a good supervisor working on optimization, Sebastian should be the first person to contact!

My first ever optimization project was an ICML paper about an asynchronous gradient method. At the time, I was quite confused by the fact that no matter what I was doing, Asynchronous gradient descent still converged. Five years later, I can finally give an answer: Because Asynchronous SGD doesn’t care about the delays, which we proved in https://arxiv.org/abs/2206.07638 our new paper. For a short summary, you can read my twitter thread about the paper or check my slides.