[RAISE at UCR] Fwd: [PDE & Applied Math seminar] Correction: Dr. Yiping Lu, Oct. 22, 10:00 AM (Pacific Time)

[Vassilis Tsotras]

Please find below information about an on-line ML-related seminar from the
Math Dept this Wednesday.

V. Tsotras


---------- Forwarded message ---------
From: <mark.alber at ucr.edu>
Date: Sun, Oct 19, 2025 at 11:49 AM
Subject: Fwd: [PDE & Applied Math seminar] Correction: Dr. Yiping Lu, Oct.
22, 10:00 AM (Pacific Time)
To: Vassilis Tsotras <vassilis.tsotras at ucr.edu>
Cc: Amit Roy Chowdhury <amitrc at ece.ucr.edu>, Mark Alber <malber at ucr.edu>


Dear Vassilis,

Could you please send the following announcement to people on the RAISE
list? Thank you.

Best regards,

Mark


---------- Forwarded message ---------
From: Yiwei Wang <yiwei.wang at ucr.edu>
Date: Fri, Oct 17, 2025 at 10:57 AM
Subject: [PDE & Applied Math seminar] Correction: Dr. Yiping Lu, Oct. 22,
10:00 AM (Pacific Time)
To: Yiwei Wang <yiweiw at ucr.edu>
Cc: Yat Tin Chow <yattinc at ucr.edu>, Yiping Lu <yiping.lu at northwestern.edu>


Dear all,

Please disregard my previous email.

The next PDE & AM seminar
<https://sites.google.com/ucr.edu/ucriverside-math-ampde-seminar/> will
take place *on Wednesday, October 22, at 10:00 AM (Pacific Time).* The talk
will be given by Dr. Yiping Lu <https://2prime.github.io/> from Northwestern
University

The seminar will be held *online only. *Please use either the link
<https://ucr.zoom.us/j/97606227247?pwd=V2d5cU1wNXV1a1NRc1N6Vm1Ja29Gdz09> or
the following Zoom ID/password:

Zoom ID: 97606227247

Passcode: 882956

The title and abstract of the talk can be found below:

*Title:* Scaling Scientific Machine Learning: Integrating Theory and
Numerics in Both Training and Inference

*Abstract: *Scaling scientific machine learning (SciML) requires overcoming
bottlenecks at both training and inference. On the training side, we study
the statistical convergence rate and limits of deep learning for solving
elliptic PDEs from random samples. While our theory predicts optimal
polynomial convergence for PINNs, optimization becomes prohibitively
ill-conditioned as networks widen. By adapting descent strategies to the
optimization geometry, we obtain scale-invariant training dynamics that
translate polynomial convergence into concrete compute and yield
compute-optimal configurations. On the inference side, I will introduce
Simulation-Calibrated SciML (SCaSML), a physics-informed post-processing
framework that improves surrogate models without retraining or fine-tuning.
By enforcing physical laws, SCaSML delivers trustworthy corrections (via
Feynman-Kac simulation) with approximate confidence intervals, achieves
faster and near-optimal convergence rates, and supports online updates for
digital twins. Together, these results integrate theory and numerics to
enable predictable, reliable scaling of SciML in both training and
inference. This is based on joint work with Lexing Ying, Jose Blanchet,
Haoxuan Chen, Zexi Fan, Youheng Zhu, Shihao Yang, Jasen Lai, Sifan Wang,
and Chunmei Wang.

For the most up-to-date list of talk announcements, please check
our website: the PDE & AM seminar’s web-site
<https://sites.google.com/ucr.edu/ucriverside-math-ampde-seminar/>.

We look forward to seeing you at the seminar.

Best regards,

Yiwei Wang and Yat Tin Chow
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ucr.edu/pipermail/raise-seminar/attachments/20251019/92afa842/attachment.htm>