Proud of my group for giving great talks at #SIAMMDS22

Congratulations to Paz and the other awardees: Abdoulaye Koroko, Sufia Khatoon, Rebekah White, Neophytos Charalambides, and Adrian De La Rocha Galan!

En route to #SIAMMDS22 in San Diego. Looking forward to a fantastic week full of exciting talks and people. Thanks to the SIAM staff for all their help putting the conference together!

We're hiring! If you (or anyone you know) wants to join a vibrant (and really nice) research community in math and data science, please consider applying at @MathEmory. More details at math.emory.edu/about/faculty…

Headed to #SIAMUQ22 in downtown Atlanta. So great to be back at a conference in person. Looking forward to meeting many old and new friends and some great talks. I’ll speak in MS49 in about 90 minutes

SIAM MDS22 is going hybrid: Speakers can choose to give their talks in-person or online. It is also not too late to propose mini-symposia, posters, contributed talks. See siam.org/conferences/CM/sd/M…

The March 1 application deadline for our REU/RET site on Computational Mathematics for Data Science is approaching. If you are (or know) an undergraduate student or in-service teacher and want to join, find more information here: math.emory.edu/site/cmds-reu…

The call for participation in SIAM's 2nd conference on Mathematics and Data Science #SIAMMDS22 is out! I'm honored and excited to co-chair this meeting. Mark your calendars for Sept 26-30,2022! siam.org/conferences/cm/conf…

If you're not already convinced that almost anything involves the SVD, check out this @TheSIAMNews article. @MathEmory undergrad Jennifer Zheng and our REU/RET student @katieekeegan explain how the SVD can help protect copyright online! sinews.siam.org/Details-Page…

Congrats to @MathEmory REU students Katy Keegan, Tanvi Vishwanath, and Yihua Xu and mentors Liz Newman and Vida John on their paper arxiv.org/pdf/2111.00587.pdf Their paper shows that treating fMRI data as tensors is more effective for classification than good old matrix-based methods

It was a lot of fun to give an interactive tag-team talk with my postdoc Liz Newman @SimonsInstitute today. We presented our joint work on exploiting deep net structure to improve training. A recording of this live experiment is available at simons.berkeley.edu/talks/ho…

Training deep neural networks can be challenging! Think learning rates, regularization, other hyperparameters... In her new paper, @MathEmory postdoc Liz Newman splits off the weights of the last layer to improve training: arxiv.org/abs/2109.14002

Congratulations to my REU/RET students Elle Buser, Emma Hart, and Ben Huenemann on finishing an exciting paper! Preprint available here arxiv.org/abs/2109.14116. Thanks also to Justin Smith for his help and @MathEmory and the @NSF for their support!

Interested in big tensor data (speech, images, videos)? Check out this paper on a rigorous mathematical framework to compress and represent your data. Congrats to my postdoc Elizabeth Newman and her collaborators on this work and its publication in PNAS pnas.org/content/118/28/e201…

I just posted my slides for tomorrow's talk in @IMAmath's Data Science Seminar. Thanks to the organizers for inviting me. I'm looking forward to presenting on recent work involving machine learning and optimal transport mathcs.emory.edu/~lruthot/sl…

Slides and codes for my short course on deep generative modeling tomorrow in the Spring School on Data and Models @usouthcarolina are now available here: mathcs.emory.edu/~lruthot/wo… This is the first iteration of the course, so I'm happy for any feedback.

Attention graduate students (and advisors)! Apply now for the GRADUATE STUDENT MATHEMATICAL MODELING CAMP @IMAmaths mathsci.udel.edu/events/conf…