Research at the Interface of Applied Mathematics and Machine Learning

CBMS Conference

Department of Mathematics, University of Houston

The Department of Mathematics at the University of Houston will be hosting the CBMS Conference: Research at the Interface of Applied Mathematics and Machine Learning from 12/08/2025 to 12/12/2025.

Schedule Overview

Time	Monday	Tuesday	Wednesday	Thursday	Friday
0800 - 0830	Check In
0830 - 0900	Welcome	Check In	Check In	Check In	Check In
0900 - 1000	Lecture 1	Lecture 3	Lecture 5	Lecture 7	Lecture 9
1000 - 1030	Refreshments	Refreshments	Refreshments	Refreshments	Refreshments
1030 - 1130	Lecture 2	Lecture 4	Lecture 6	Lecture 8	Lecture 10
1130 - 1300	Lunch	Lunch (**)	Lunch (##)	Lunch	Lunch (##)
1300 - 1345	Talk	Talk	Talk	Talk
1345 - 1430	Talk	Talk	Talk	Talk
1430 - 1500	Refreshments	Refreshments	Refreshments	Refreshments
1500 - 1545	Talk	Panel	Discussion	Talk
1545 - 1630	Talk	Panel	Discussion	Talk
1630 - 1700	Break			Social Event
1700 - 1900	Poster			Social Event

(**) Group Picture; (##) Lunch will be provided.

Check in: Pick up your badge in SEC101 of the Science and Engineering Classrooms

Poster Session Venue: Ballroom of Student Center South

Conference Venue: Room SEC101 of the Science and Engineering Classrooms

Food: We offer refreshments twice a day. For lunch the participants are on their own. We provide on campus food options on the Venue webpage. There is one exception: We will provide lunch boxes on Friday.

Talks

We have scheduled 12 talks from leading authorities in (Sci)ML from Monday 12/08 through Thursday 12/11.

Monday, 12/08/2025

1300 - 1345 Demetrio Labate, Mathematics, University of Houston
Title: Optimal approximations of multivariate functions on smooth manifolds using deep ReLU neural networks
Show Abstract
The expressive power of deep neural networks is manifested by their remarkable ability to approximate multivariate functions in a way that appears to overcome the curse of dimensionality. This ability is exemplified by their success in solving high-dimensional problems where traditional numerical solvers fail due to their limitations in accurately representing high-dimensional structures. Under the assumption that a function is defined in a $d$-dimensional manifold embedded in a space of higher dimension $D$, we prove that the uniform convergence estimates of the approximation by deep neural networks with ReLU activation functions do not depend on the ambient dimension $D$ but only on its lower manifold dimension $d$, in a precise sense. Our result improves existing results from the literature where the approximation was shown to depend weakly on $D$.
This is joint work with Ji Shi (University of Houston)
1345 - 1430 Aurya Javeed, Optimization and Uncertainty Quantification, Sandia National Laboratories
Title: ProxSTORM: A Stochastic Trust Region Algorithm for Nonsmooth Optimization
Show Abstract
We present a stochastic trust-region algorithm for minimizing the sum of a nonconvex Lipschitz-smooth function and a nonsmooth convex function evaluated exactly. This algorithm, which we call ProxSTORM, generalizes STORM---a stochastic trust-region algorithm for the unconstrained optimization of smooth functions---and an inexact deterministic proximal trust-region algorithm. Our problem assumptions are essentially those of STORM when the convex function is zero. Our analysis follows STORM in using martingale arguments to prove global convergence and an expected complexity bound. We apply ProxSTORM to $\ell^1$-regularized neural network training and to topology optimization.
This is joint work with Robert J. Baraldi (Sandia), Drew P. Kouri (Sandia), and Katya Scheinberg (Georgia Institute of Technology).
1500 - 1545 Jonathan Siegel, Mathematics, Texas A&M
Title: Universal Approximation for In-Context Operator Learning
Show Abstract
In this talk, we consider the problem of in-context learning of operators using neural networks. In this paradigm, a neural network is initially trained on a range of operators. At inference time, the network is passed a sequence of example inputs and outputs, and based upon this it infers the operator and can apply it to a new input. We will give a precise mathematical formulation of this problem, and show that a newly developed architecture, DeepOSets, is universal in the sense that it can in principle learn any compact collection of operators.
This is joint work with Steven Chiu, Aditya Nambiar, Ali Hamza Abidi Syed, and Ulisses M. Braga-Neto (all from Texas A&M)
1545 - 1630 Lu Zhang, CMOR, Rice University
Title: Data-driven hyperbolic conservation laws
Show Abstract
Hyperbolic conservation laws are fundamental to modeling wave propagation, fluid dynamics, and collective behavior across science and engineering. Despite their universal importance, traditional numerical solvers rely on explicitly known flux functions and parameters, which are often unavailable in realistic scenarios where only trajectory or observation data are accessible. In this talk, I will discuss our recent work in data-driven approaches for learning hyperbolic conservation laws that preserve their intrinsic physical and mathematical structures. These developments combine principles from numerical analysis, partial differential equations, and machine learning to construct models that are both predictive and faithful to the underlying conservation and stability properties of the governing equations.
This is joint work with Lizuo Liu (Dartmouth College) and Anne Gelb (Dartmouth College)

Tuesday, 12/09/2025

1300 - 1345 Wenjing Liao, Mathematics, Georgia Tech
Title: Exploiting Low-Dimensional Data Structures and Understanding Neural Scaling Laws of Transformers
Show Abstract
When training deep neural networks, a model’s generalization error is often observed to follow a power scaling law dependent on the model size and the data size. A prominent example is transformer-based large language models (LLMs), where networks with billions of parameters are trained on trillions of tokens. A theoretical interest in LLMs is to understand why transformer scaling laws emerge. In this talk, we exploit low-dimensional structures in language datasets by estimating its intrinsic dimension and establish statistical estimation and mathematical approximation theories for transformers to predict the scaling laws. This perspective shows that transformer scaling laws can be explained in a manner consistent with the underlying data geometry. We further validate our theory with empirical observations of LLMs and find strong agreement between the observed empirical scaling laws and our theoretical predictions. Finally, we turn to in-context learning, analyzing its scaling behavior by uncovering a connection between the attention mechanism in transformers and classical kernel methods in machine learning.
This is joint work with Alex Havrilla (Georgia Institute of Technology), Zhaiming Shen (Georgia Institute of Technology), Alex Hsu (Purdue University), and Rongjie Lai (Purdue University)
1345 - 1430 Simon Foucart, Mathematics, Texas A&M
Title: Worst-Case Learning: the View from Optimal Recovery
Show Abstract
Rooted in Approximation Theory, Optimal Recovery can be viewed as a trustworthy learning theory focusing on the worst case.
Regrettably, compared to more popular Machine Learning alternatives, the classical theory of Optimal Recovery overlooked the computational aspect, with a few exceptions, e.g. the development of spline functions.
Nowadays, modern optimization techniques facilitate advances---even theoretical ones---on the minimax problems that abound in the field.
The talk will illustrate this point by showcasing a few snippets from the speaker's recent work (e.g. full recovery from deterministically inaccurate data in Hilbert spaces, prediction of vector-valued functions based on merely convex models, prediction of the maxima of Lipschitz functions from inaccurate point values).

Wednesday, 12/10/2025

1300 - 1345 Raghu Bollapragada, Operations Research and Industrial Engineering, The University of Texas at Austin
Title: Adaptive, Fast, and Scalable Methods for Large-Scale Nonlinear Optimization
Show Abstract
Nonlinear optimization problems arise in a wide range of applications, from inverse problems to training deep learning models. The scale, computational cost, and complexity of these models make classical optimization techniques impractical. To address these challenges, we propose new optimization methods that are particularly well-suited for distributed computing implementations. Our techniques employ adaptive sampling strategies that gradually increase the accuracy of step computations to enhance efficiency and scalability. Additionally, they incorporate second-order information through Hessian averaging approaches, resulting in faster convergence. We provide some interesting (and perhaps surprising) fast local convergence results for our methods. The effectiveness of the proposed algorithms is demonstrated through numerical experiments on large-scale machine learning models.
This is joint work with Chengyue Zhang (The University of Texas at Austin) and Thomas O'Leary-Roseberry (Ohio State University)
1345 - 1430 Deepanshu Verma, Mathematical & Statistical Sciences, Clemson University
Title: Neural Network Approaches for Optimal Control: Implicit Hamiltonians and Transferable Policies

Show Abstract
This talk presents two neural network methodologies advancing optimal control beyond current limitations. First, we address implicit Hamiltonians in practical problems like space shuttle reentry, where existing methods fail without explicit feedback control formulas. Our end-to-end implicit deep learning approach directly parameterizes value functions to handle the underlying implicit structure while enforcing physical principles through the relationship between optimal control and value function gradients, bridging Pontryagin's Maximum Principle and Dynamic Programming. Using Jacobian-free backpropagation, we efficiently train implicit networks for high-dimensional feedback controllers in previously intractable scenarios.
Second, we tackle the computational burden of re-solving problems when objectives change. Our function encoder framework learns reusable neural basis functions enabling zero-shot adaptation through offline-online decomposition: basis functions are learned once, while adaptation requires only lightweight coefficient estimation. Experiments demonstrate near-optimal performance across diverse dynamics with minimal overhead.
These approaches expand neural HJB applicability by handling structural complexity through implicit Hamiltonians and enabling operational flexibility through transferable policies for real-time deployment.
This is joint work with Samy Wu Fung (Colorado School of Mines), Xingjian Li (UT Austin), Krishna Kumar (UT Austin), Nicole T. Yang (University of Tennessee), Kelvin Kan (UCLA), Ján Drgoňa (JHU), Stanley Osher (UCLA), Eric Gelphman (Colorado School of Mines)

Thursday, 12/11/2025

1300 - 1345 Eric C. Chi, Statistics, University of Minnesota
Title: Optimization for Sampling and Sampling for Optimization
Show Abstract
Sampling and optimization are two general computational approaches that enable statistical modeling and machine learning. This talk will highlight recent work on how they can help each other in practice. The first part of the talk covers how algorithmic primitives from non-smooth optimization can aid Markov chain Monte Carlo (MCMC). We discuss how the Moreau-Yosida envelope can be used to make importance sampling more reliable for estimating characteristics of a high-dimensional target distribution. The second part of the talk covers the utility of Monte Carlo approximations of proximal maps within widely used fixed-point methods.
The first part is joint work with Apratim Shukla (IIT Kanpur) and Dootika Vats (IIT Kanpur). The second part is joint work with Nick Di (Rice University) and Samy Wu Fung (Colorado School of Mines).
1345 - 1430 Akil Narayan, Mathematics, University of Utah
Title: Certifiable operator learning from data: Building approximations to PDE solution maps
Show Abstract
We consider the problem of learning an unknown, possibly non-linear operator $K:\mathcal X \to \mathcal Y$, with $\mathcal X, \mathcal Y$ separable Hilbert spaces. This is accomplished via a supervised learning procedure, with independent samples $f^i$ drawn from a tailored probability measure on $\mathcal{X}$ and (possibly noisy) observations $g^i=K(f^i) + \eta^i$. For a fixed probability measure $\rho$ on $\mathcal{X}$, the admissible class of operators, $L^2_\rho(\mathcal X;\mathcal Y)$, is taken to be the Hilbert space of Bochner square-integrable maps, and the reconstruction error is measured in this norm. Given an $N$ dimensional subspace $V$ of $L^2_\rho(\mathcal X;\mathcal Y)$, we establish probabilistic accuracy and stability results for general weighted least squares approximations in $V$, and we show that there exists a sampling measure $\mu$ and weight functional $w$ for which near-optimal stability and accuracy can be achieved for a near-minimal sample size. We provide explicit procedures for constructing approximation spaces and sampling from the associated optimal measures in cases of interest. Finally, we highlight the effectiveness of this method in several numerical experiments: these include learning the PDE solution operator to the Poisson, Navier-Stokes, and viscous Burgers equations.
This is joint work with John Turnage (University of Utah), Matthew Lowery (University of Utah), John D. Jakeman (Sandia National Laboratories), Zachary Morrow (Sandia National Laboratories), and Varun Shankar (University of Utah)
1500 - 1545 Juntao Huang, Mathematics & Statistics, University of Delaware
Title: Hyperbolic machine learning moment closure models for kinetic equations
Show Abstract
In this talk, we take a data-driven approach and apply machine learning to the moment closure problem for the kinetic equations, including radiative transfer equation and Boltzmann BGK equations. Traditional closures often rely on empirical assumptions, while naive machine learning closures can violate structural properties of the PDEs, leading to ill-posedness and numerical instability. To address these challenges, we propose a gradient-based moment closure, where neural networks directly learn the gradient of the high-order moment. Furthermore, we develop two strategies to enforce hyperbolicity, ensuring well-posed and stable evolution.Extensive benchmark tests—including variable-scattering, Gaussian-source, and two-material problems—demonstrate that our hyperbolic ML closures achieve high accuracy, robust stability, and strong generalization beyond training regimes.
1545 - 1630 Elizabeth Newman, Mathematics, Tufts Univeristy
Title: Exploiting Structure for Reliable, Efficient Training
Show Abstract
Deep neural networks (DNNs) have been successful high-dimensional function approximators in countless applications. However, training DNNs is notoriously challenging, requiring significant time and computational resources. In this talk, we will make training easier by exploiting commonly used network structures, incorporating second-order information, and ensembling to reduce computational overhead and improve generalization. In the first part of the talk, we will describe training strategies designed for separable DNNs in which the weights of the final layer are applied linearly. We will leverage this linearity in two ways: using partial optimization (variable projection) in the deterministic setting and iterative sampling in the stochastic setting. We will demonstrate empirically that both approaches yield faster convergence to more accurate DNN models and less sensitivity to training hyperparameters. In the second part of the talk, we will further leverage variable projection (VarPro) to enhance gradient boosting, a greedy training strategy that iteratively optimizes lightweight learners to build an expressive ensemble. We will discuss some theoretical benefits of VarPro in the boosting setting and provide empirical support of the benefits of boosting with VarPro on a variety of regression and classification tasks.

Lectures

The workshop features ten lectures within 3 modules. An outline can be found below. The video recording for these lectures can be found here: lecture recordings.

Module 1: Machine Learning Crash Course (3 Lectures)

This module aims to define ML techniques and commonly used terms mathematically to make them accessible to applied and computational mathematicians.

Show Details

Lecture 1: Overview

This lecture sets the main notation for this conference and discusses different forms of ML that are relevant to this workshop. For simplicity, we define learning as designing a function or operator $f_\theta$ and learning its weights $\theta$ to accomplish a given task. In this workshop, we focus on $f_\theta$ as a neural network. This lecture overviews terminology and different learning tasks in applied mathematics. The remaining lectures in this module discuss the two crucial aspects of learning in more detail. To give participants a broad context, we provide motivating examples from unsupervised, semi-supervised, supervised learning, generative modeling, reinforcement learning, and operator learning. Where possible, we link the examples to corresponding problems in applied mathematics. For example, supervised learning corresponds to data fitting, generative modeling has links to optimal transport, reinforcement learning is tied to optimal control, and operator learning arises in solving PDEs and inverse problems.

Lecture 2: Neural Networks

This lecture is devoted to different ways to design the neural network architecture that defines $f_\theta$. This is a crucial step in solving practical problems. We review classical multilayer perceptron architectures, where $f_{\theta}$ is defined by concatenating affine transformations and pointwise nonlinear activation functions. While these architectures are universal function approximators and can be effective in many applications, they have difficulties approximating simple functions, like the identity mapping, and their training is challenging with increasing depth. Adding skip connections, as done in residual networks, can overcome this disadvantage and be trained with hundreds or thousands of layers. The latter architectures provide links to optimal control and will be revisited in the next module. We also present graph neural networks, which are crucial to handling unstructured data, and give a mathematical description of transformers, including their attention mechanism.

Lecture 3: The Learning Problem

This lecture introduces the loss functions that can be used to train the neural network architectures for a given task. For supervised learning, we discuss regression and cross-entropy losses. We discuss maximum likelihood training and variational inference with the empirical lower bound (ELBO) for generative modeling. As an example of unsupervised learning, we discuss PDE losses in PINNs. We illustrate the difference between minimizing the loss function and learning, which requires generalization to unseen data, using examples from polynomial data fitting that most participants will recall. We then provide further insights by discussing the integration error of Monte Carlo approximation of the loss.

Module 2: Applied Mathematics for Machine Learning (3 Lectures)

This module discusses three themes of applied mathematics research in ML. We spend one lecture per area and aim to introduce the core principles that underlie recent advances we expect to be discussed by several invited presenters.

Show Details

Lecture 4: Stochastic Optimization

Having defined the ML model and loss functions, this lecture discusses optimization algorithms that can be used to identify weights. Since the loss functions usually involve high-dimensional integrals, we approximate them using Monte Carlo integration. This naturally leads to stochastic optimization algorithms such as stochastic gradient descent and its variants. In this lecture, we discuss convergence properties, theoretical and empirical results that show convergence to global minimizers for highly nonconvex functions, and their ability to regularize the problem.

Lecture 5: Regularization

This lecture investigates the relation between generalization and regularization in more depth. Building upon advances in applied mathematics, we discuss iterative regularization, direct regularization, and hybrid approaches in the context of ill-posed inverse problems. From this perspective, we show new insights into the double-descent phenomenon arising in modern ML. We illustrate this using the random feature models and demonstrate that adequate regularization can help those models generalize to unseen data. We also review recent results on the benefits and challenges of adding regularization theory into stochastic optimization schemes.

Lecture 6: Continuous-in-Time Architectures

This lecture surveys neural network architectures whose depth corresponds to artificial time and whose forward propagation is given by differential equations. As a starting point, we view the infinite-depth limit residual networks as forward Euler discretizations of a nonstationary, nonlinear initial value problem. We discuss the theoretical implications for supervised learning and the opportunities in generative modeling. We present extensions to PDEs when the features correspond to image, speech, or video data and the benefits of generalizing the framework to stochastic dynamics. The latter allows us to discuss image-generation algorithms like Dalle-2 and other techniques based on score-based diffusion.

Module 3: Machine Learning for Applied Mathematics (4 Lectures)

This module discusses four avenues of applying ML techniques in applied mathematics problems.

Show Details

Lecture 7: Scientific Machine Learning

This lecture demonstrates the use cases and challenges of using ML in scientific contexts. Unlike traditional big-data applications, specific challenges arise from scarce datasets, the complex nature of observations, the lack of similar experience, which makes finding hyperparameters difficult, and so on. We also overview important research directions not covered in the remaining lectures, such as PINNs, neural operator learning, and reinforcement learning.

Lecture 8: High-dimensional PDEs

This lecture introduces several high-dimensional PDE problems for which ML techniques provide promising avenues. While many PDE problems are phrased in physical coordinates and are thus limited to three dimensions plus time, there are several areas in which the dimensions become considerably higher, and the curse of dimensionality limits traditional numerical techniques. We use examples including the Black Scholes equations in finance and Hamilton Jacobi Bellman equations in optimal control. In those cases, the curse of dimensionality can be mitigated (but not entirely overcome) by appropriate numerical integration, adequate neural network models, and training.

Lecture 9: Inverse Problems

Due to their ill-posedness and (in some cases) abundance of training data, applying ML techniques to improve the solution of inverse problems has been a promising research direction. In this lecture, we will first demonstrate that approximating the inverse map with a neural network in a supervised way in ill-posed inverse problems leads to unstable approximations. We then consider Bayesian formulations of the inverse problem as they rely on modeling and analyzing complex, high-dimensional probability distributions. This provides links to generative modeling, and we show how those techniques can overcome the limitations of the traditional Bayesian setting and improve computational efficiency.

Lecture 10: Mathematical Reasoning and Discovery

This lecture highlights recent trends in mathematics that employ large language models ({\bf LLM}s) to discover and reason about mathematics. For example, combining proof assistants with LLMs assists in formalizing mathematical theorems, proofs, and even sketch proofs. Another example is the use of reinforcement learning to discover counter-examples in combinatorics. We discuss funsearch and its ability to develop interpretable solutions to challenging discrete math problems. In the context of computational mathematics, reinforcement learning is noteworthy for discovering more efficient implementations of matrix-matrix products.

Panel

The conference will host a panel where experts and scientists share experiences while discussing current and future trends in research and machine learning applications. The panel discussion is scheduled for Tuesday 12/9, 1500 to 1630. Invited panelists include:

Detlef Hohl, Shell Technology Center Houston

Dr. Hohl spent his entire Shell career in R&D, both in the US and in Europe. First in geophysical seismic imaging, then in probabilistic seismic inversion, and then he became R&D team leader for “Quantitative Reservoir Management”. From 2010-2017 he was R&D General Manager Computation and Modeling where he led a project portfolio in data analytics, computational engineering and materials science, geoscience and petroleum engineering.

Dr. Hohl was appointed Shell’s Chief Scientist for Computational and Data Science in 2017 where he oversees and guides Shell’s entire computational and computer science portfolio, including elements of Artificial Intelligence, physical systems simulation at all spatial and temporal scales, chemicals and chemical engineering modeling, future energy systems optimization, atmospheric and Earth science modeling.

Dr. Hohl has always used the largest available high-performance computers of their time to do big things that cannot be done otherwise. He is active in the academic, National laboratory and joint industry research communities, member of APS, ACM, SIAM, SPE, SEG and AGU. Dr. Hohl is adjunct professor and teaches courses at Rice University (Computational and Applied Math). He held various temporary and visiting positions at NCSA, SISSA Trieste, NIST and Stanford University. In his free time, science remains his biggest hobby.

Xiaoqian Jiang, School of Biomedical Informatics, UTHealth Houston

Dr. Xiaoqian Jiang is the Associate Vice President for Medical AI, Chair of the Department of Health Data Science and Artificial Intelligence, and the Christopher Sarofim Professor at The University of Texas Health Science Center at Houston (UTHealth). He also directs the Center for Secure Artificial Intelligence for Healthcare (SAFE) at McWilliams School of Biomedical Informatics. Dr. Jiang is a leading expert in privacy-preserving data mining, federated learning, and explainable machine learning, with over $35 million in grant funding from NIH and other prestigious awards such as the CPRIT Rising Stars and UT Stars. His research spans a range of critical health AI applications, from Alzheimer’s prevention to COVID-19 patient tracking, and his innovative work in human-in-the-loop AI models and computational phenotyping has earned several best paper awards from AMIA. Dr. Jiang's mission aligns with advancing healthcare AI, fostering high-quality education, and improving patient outcomes through AI-driven innovations.

Peter Kochunov, Psychiatry and Behavioral Sciences, UTHealth Houston

Dr. Kochunov is a board-certified MRI physicist with over two decades of experiences in development of novel data analysis protocols with emphasis on the quantitative, multimodal analyses of genetic factors that are responsible for structural and functional variability.

Dr. Kochunov has a background in neuroimaging, electrical engineering, software development and statistics. Dr. Kochunov has participated in development of many popular neuroimaging tools and formats including SOLAR-Eclipse, ENIGMA-Viewer, ENIGMA-DTI and ENIGMA-rsFMRI analyses pipelines, Talairach deamon, BrainMap, Mango and BrainVisa Morphologist, NFITI and others.

Dr. Kochunov’s research is described in over 300 scientific manuscripts, including some of the first manuscripts on heritability of white matter integrity, gray matter thickness, resting-state connectivity and application of these approaches in severe mental illness research.

Javad Razjouyan, Baylor College of Medicine

Javad Razjouyan, Ph.D., is an Assistant Professor (tenure track) of medicine for Health Services Research and the Institute for Clinical & Translational Research (ICTR) at Baylor College of Medicine (BCM). He is also a health research scientist at Implementation Science & Innovation Core, Center for Innovations in Quality, Effectiveness and Safety (IQuESt) at Michael E. DeBakey VA Medical Center. He is faculty member of the Big Data-Scientist Training Enhancement Program (BD-STEP) of the Department of Veterans Affairs (VA) and the National Cancer Institute (NCI) and he is adjunct Assistant Professor of Epidemiology at UTHealth Houston School of Public Health.

Dr. Razjouyan serving as the Director of Artificial Intelligence in Health Lab (AIH-Lab) at BCM and co-director of BD-STEP advanced fellowship program, Houston site.

He has published more than 40 scientific publications in peer-reviewed journals, more than 50 conference proceedings or abstracts, three filled patents, and a textbook for undergraduate students in biomedical engineering. He received a young investigator award from the Gerontological Society of America conference in 2014 as digital biomarker development. He also won a junior investigator travel award at the American Heart Association - Quality of Care & Outcome Research 2019 as he developed an EMR based frailty index by machine learning techniques. He mentored three advanced fellows at the BD-STEP program on used of EMR and performing advanced machine learning algorithms. Currently, He is mentoring three advanced post-doctoral fellows at the AIH-lab. His post-doctoral fellows use artificial intelligence (AI), machine learning (ML) algorithms, and natural language processing (NLP) tools on various medical fields such as sleep medicine, psychology, dementia, heart failure, and frailty.

Amir Sharafkhaneh, Baylor College of Medicine

Dr. Amir Sharafkhaneh is a Professor of Medicine (tenured) at Baylor College of Medicine and a leading authority in sleep medicine. He completed his medical degree at Tehran University of Medical Sciences, followed by an internal medicine residency at Long Island College Hospital and a fellowship in Pulmonary, Critical Care, and Sleep Medicine at Baylor College of Medicine, where he also earned a PhD in medical research.

With over 25 years of clinical and academic experience, Dr. Sharafkhaneh has authored numerous peer-reviewed publications and book chapters in the fields of pulmonary and sleep medicine. He founded the first accredited sleep medicine fellowship program in Texas and has since trained more than 100 sleep specialists. His work has been supported by multiple federal grants, including initiatives to develop telemedicine programs that expand access to sleep care in underserved areas.

Dr. Sharafkhaneh currently co-chairs the VA clinical practice guideline committees for obstructive sleep apnea, insomnia, asthma, and COPD. His research team applies artificial intelligence and advanced data analytics to large-scale electronic health record data to advance the understanding and treatment of sleep and respiratory disorders. He also co-leads the AI Interest Group of the World Sleep Society.

Xiao-Hui Wu, ExxonMobil Technology & Engineering

Xiao-Hui Wu joined ExxonMobil Upstream Research Company in 1997. His research experience covers geologic modeling, unstructured gridding, upscaling, reduced order modeling, and uncertainty quantification. He is a Senior Earth Modeling Advisor in the Computational Science Function. Xiao-Hui received his Ph.D. in Mechanical Engineering from the University of Tennessee and worked as a postdoc in Applied Mathematics at Caltech before joining Exxon Mobil. He is a member of SPE and SIAM, a technical editor/reviewer for the SPE Journal, Journal of Computational Physics, and Multiscale Modeling and Simulation. He served on program committees of several conferences, including the Reservoir Simulation Symposium.

Poster Session

We have scheduled a poster session on Monday 12/08, 5 PM to 7 PM in the Ballroom of the Student Center South.

The poster boards are of size (height x width) = (36" x 48") (landscape format). Please plan accordingly.

Show Details

Bounding the Generalization Error of ReLU Networks (Tom Winckelman, Texas A&M University)
A Unified Neural Network Framework for Efficiently Solving Ordinary Differential Equations (Pavithra Venkatachalapathy, Texas Tech University)
Mixed Finite Element Methods for Boundary Control Problems Constrained by the Biharmonic Equation (Karthik Kumar Vasudeva, Texas Tech University)
Solving Nonlinear PDEs with Sparse Radial Basis Function Networks (Zihan Shao, University of California)
Graph-Based Multi-View Learning: Joint Optimization of Structure Preservation, Feature Selection, and Regularization (Wanjun Ning, The University of Texas at Arlington)
Streamflow-Informed Correction of Radar Rainfall Biases for Improved Flood Forecasting (Zhihua Li, University of Iowa)
Accelerated Optimization in ML: Challenges of Non-Convexity and Noisy Gradients (Kanan Gupta, University of Pittsburgh)
Machine Learning--Driven Mitral Valve Detection in Echocardiography via Morphological Dilation and Classification (Ahmed Al Hasan, University of Houston)
Matrix analysis for shallow ReLU neural network least-squares approximations (Tong Ding, Purdue University)
In-Context Operator Learning for Parametric PDEs with DeepOSets (Shao-Ting Chiu, Texas A&M University)
Estimating Manifold's Extrinsic Curvatures by Closest Point Projection of Rays (Liangchen Liu, The University of Texas at Austin)
Superfast and stable divide-and-conquer singular value decomposition for hierarchical rank-structured matrices (Chenyang Cao, Purdue University)
Deep Learning Outperforms Traditional Machine Learning Method in Predicting Childhood Malnutrition: Evidence from Survey Data (Deepak Bastola, Florida Atlantic University)
Bayesian Inference on SPD Manifolds: Geometry-Aware Learning of Posterior Covariances (Pegah Amiri, University of Houston)
Belief Based Decision Making Under Uncertainty: A Computational and Empirical Analysis of Optimal and Human Strategies (Parnian Ahmadzadeh, University of Houston)
Tensorial Reduced-Order Methods for Multispecies Tumor Growth Simulations (Asikul Islam, University of Houston)
Learning Ecological and Epidemic Processes (Hyangim Ji, Texas A&M)
ReBaNO: Reduced Basis Neural Operator Mitigating Generalization Gaps and Achieving Discretization Invariance (Haolan Zheng, University of Massachusetts Dartmouth)
Squared Wasserstein Distance for Efficient Reconstruction of Stochastic Differential Equations (Mingtao Xia, University of Houston)
Transformer–LSTM with Multitask Quantile Learning for Bitcoin Volatility and VaR Forecasting (Emmanuel Tomiwa Siyanbola, Western Kentucky University)
Nonlinear functional regression by functional deep neural network with kernel embedding (Zhongjie Shi, Georgia Institute of Technology)
Understanding In-Context Learning on Structured Manifolds: Bridging Attention to Kernel Methods (Zhaiming Shen, GEorgia Institute of Technology)
Can Classical Models and Machine Learning Work Together to Predict Cancer Cell Growth? A Comparative Study of Differential Equation-Based Models and Physics-Informed Neural Networks for Biomedical Forecasting (Widodo Samyono, Jarvis Christian University)
Topological database for global dynamics of ODEs (Bernardo Rivas, University of Toledo)
Toward Realistic RNA-Seq Simulations: A Comparative Evaluation of Current Approaches (Nazia Riasat, North Dakota State University)
Physics Informed Neural Networks for a Wigner-Fokker-Planck Model of Open Quantum Systems (Jose Morales Escalante, University of Texas at San Antonio)
Application of Machine learning for sub-grid flux Parametrization in the Shallow water equations (Md Amran Hossan Mojamder, University of Houston)
Sampling with Bayes Hilbert splines (S. David Mis, Rice University)
Data Driven Learning to Enhance a Kinetic Model of Distressed Crowd Dynamics (Kamrun Nahra Mily, University of Houston)
Zero-Shot Transferable Solution Method for Parametric Optimal Control Problems (Xinjian Li, The University of Texas at Austin)
Understanding Neural Collapse in ReLU Networks through Truncation Maps and Cone Geometry (Mayank Konduri, The University of Texas at Austin)
A Feature-Based Approach to Shape Graph Analysis (Murad Hossen, University of Houston)
A Convergent Generalized Krylov Subspace Method for Compressed Sensing MRI Reconstruction with Gradient-Driven Denoisers (Tao Hong, The University of Texas at Austin)
Towards Diffusion-Based Trajectory Optimization for Singular Control Problems (Samuel Glickman, University of Hawaii at Manoa)
A Support Vector Machine Based Artificial Intelligence Technique Using Genetic Algorithms to Screen Metabolites Associated with Heart Disease in the Qatari Population (Ryad Ghanam, Virginia Commonwealth University in Qatar)
A Global–Local Multiscale Markov Chain Monte Carlo Method for Domain-Decomposed Bayesian Inference (Aidan Gettemy, University of Texas at Dallas)
Learning Low-Dimensional Nonlinear Dynamics with Windowed Autoencoder Networks (Biraj Dahal, Georgia Tech)
Multi-Armed Bandit with Kaplan-Meier Estimation of Long-Term Trials, Yangfan Cui, North Carolina State University)
Boosting VarProNets: Efficient Gradient Boosting with Variable Projection (Abhijit Chowdhary, Tufts University)
Optimal Sensor Placement for Gaussian Processes using Column Subset Selection (Jessie Chen, North Carolina State University)
Polyharmonic Splines for Operator Learning (Desmond Boateng, Boise State University)
Mathematics-Guided Deep Learning Framework for Genomic Sequence Classification in Liver Cancer (Sanaa Anjum, University of Wah)
Adaptive Data-Driven Reduced-Order Models for inverse scattering problems in lossless and lossy media (Anarzhan Abilgazy, Southern Methodist University)
Interpretable Learning in Hamiltonian and Dispersive Wave Systems (Jimmie Adriazola, Arizona State University)
Tensorial Reduced-Order Methods for Multispecies Tumor Growth Simulations (Asikul Islam, University of Houston)
In-Context Operator Learning for Parametric PDEs with DeepOSets (Shao-Ting Chiu, Texas A&M University)