Lionel Tondji

Super excited to tell you that I've joined @Dirque_L 's team as an ESR and I will be working on a @opt_trade project at @tuBraunschweig. Can't wait to start this new chapter in my life 😀.

Job Alert: Are you thinking of doing a PhD in ML, with a theoretical/algorithmic flavor? I'm hiring talented and passionate students to work with me on: ML theory; Neural scaling laws; Synthetic data (the good, the bad and the ugly); Explainable and trustworth AI (adversarial robustness, alignment, algorithmic fairness, etc.); New models and paradigms Location. Concordia University and Mila Intitute, Montreal, Canada. On-site. Source of funding: Public (CRC grant, etc.) Requirements - Hold a MSc degree (or equivalent) in CSC, Math, Statistics, or Physics. Very strong BSc holders will be considered -A Solid math background. - ML fundaments. Familiarity with Numpy/Scipy/Matplotlib + sklearn/Pytorch (or equivalent). -ML theoryed papers in recognizable ML conferences (neurips/iclr/icml/aistats/aaai/colt, etc.) is a plus.- Contact. DM or find my email on my website hosted at Concordia University. Message should have informative title. You should have some familiarity with my work and should clearly indicate why you think you're good a match for this job posting. I won't respond to requests which appear to be "at random".

I have an open postdoc (or phd) position! Project concerns PINNs and material science. Please reach out if you have any questions! (Job ad only in German, unfortunately…) uni-bremen.de/en/university/…

You have a math degree and are interested in climate research for a PhD? 🧮🧊🌏 Then this job ad is for you! Supervision will be by Monica Ionita from the AWI and myself from the Center for Industrial Math. jobs.awi.de/Vacancies/1771…

I am eager to share my knowledge and experiences with the next generation of AI pioneers and to be part of this transformative journey once again. #AMMI #MachineLearning #AI #Education #Africa #Mbour #Teaching #GivingBack (3/3)

AMMI is a fully-funded, intensive graduate program that provides brilliant young Africans with state-of-the-art training in machine learning and its applications. aimsammi.org (2/3)

I am thrilled to announce that I am currently in Mbour, Senegal to teach a course at the AMMI program. It feels like coming full circle, as I once walked these halls as a student and now have the incredible opportunity to contribute as an instructor. (1/3)

@MathsCirclesOz Algebraic Geometry : Lie Algebras

What would you put in the middle?

@Jony_Chinijo Can we say it is an anti equivalence relation?

Can we say it is an anti-distance?

Let X be the space of humans and define the relation d(x, y) = friendship between x and y. Now, check if it is a distance. It does not satisfy any of them! - Im not necessarily friend of myself, - If I am your friend, you are not my friend - Im not the friend of your friends.

@A__Diack This work wouldn't have been possible without the incredible efforts of @Idriss_Tondji and @Dirque_L

@professordesJ Cc: @Idriss_Tondji , @Dirque_L

🔥New paper alert!🔥Do you want to solve linear inverse problems with noisy measurements? We have a solution for you! Our new paper introduces a Bregman-Kaczmarz method with an adaptive stepsize that can recover the exact solution. Check it out! (with @Idriss_Tondji & @Dirque_L)

Requiring a publication in NeurIPS or the like before applying to PhD is toxic. It drives subpar submission to conferences. It selects privileged background. Senior Academics, please fight this, for the sake of the community

Any amount will be helpful.

New preprint 📄 "Linearly convergent adjoint free solution of least squares problems by random descent" by Felix Schneppe, @ltondji and myself arxiv.org/abs/2306.01946 In this paper we revisited an old problem: What if you want to minimize ||Av-b||² 𝗕𝗨𝗧 1/10

@bremen79 Nice post, I really enjoyed reading it and I learned a lot although I am not familiar with online learning. Cor3 looks like a deterministic general version of the expected Smoothness By Gower et al.

🚨New Blog Post🚨 Adapting to Smoothness with Normalized Gradients A "new" result:normalized gradients give adaptation to local Hölder smoothness,for free! The trick is general:any first-order optimization algorithm can be used, even parameter-free ones parameterfree.com/2023/06/19/ada…

Want to get up to speed with optimal transport? Paul Manns, Christian Meyer and I have the right summer school for you. Learn from the best speakers: @NenLuke, FX Vialard, and Bernhard Schmitzer Register here: mathematik.tu-dortmund.de/lsx/cms/de/spp…

Here are the links: meetings.siam.org/sess/dsp_progr… meetings.siam.org/sess/dsp_progr… meetings.siam.org/sess/dsp_progr…

At #SIAMOP23 and interested in randomized methods for linear systems? Still here on Saturday? Yes to both questions? Then our minisymposium "Randomized Methods for Linear Systems: Row-Action and Coordinate Descent" is just for you! Jointly organized with @ltondji

@Dirque_L n² - n = n(n-1) which is always even.

I raise the bar and ask "If n is a number, which one of the following is *always* and even number?" - n² - n²-2 - (n-2)2 - n²-n