Most existing AI learning methods can be categorized into supervised, semi-supervised, and unsupervised methods. These approaches rely on defining empirical risks or losses on the provided labeled and/or unlabeled data.

Hajj is a very complex event that requires huge logistical organization. Mobility is one of the most important logistical services needed since Hajj rituals are based on movement from one location to another.

Quadratic Forms in Gaussian Random Variables (QFGRV) appear in a wide variety of applications in signal processing, communications, performance and reliability engineering, genetics, etc.

Analogies between financial markets and critical phenomena have long been observed empirically. So far, no convincing theory has emerged that can explain these empirical observations. Here, we take a step towards such a theory by modeling financial markets as a lattice gas.

Living in a computer-reliant era, we’re balancing the power of computer systems with the challenges of ensuring their functional correctness and security. Program analysis has proven successful in addressing these issues by predicting the behavior of a system when executed.

The research focuses on improving metal-GaN and dielectric-GaN interfaces for high-performance GaN-based electronics. For metal-GaN, the damage caused by e-beam evaporation was mitigated using Ti3C2Tx MXene films, achieving a record ION/IOFF current of 1013 and low subthreshold swing.

Abstract

Neural networks have achieved impressive performance in many applications such as image and speech recognition and generation. State-of-the-art performance is usually achieved via a series of engineered modifications to existing neural architectures and their training procedures. However, a common feature of these systems is their large-scale nature: modern neural networks usually contain Billions - if not 10's of Billions - of trainable parameters, and empirical evaluations (generally) support the claim that increasing the scale of neural networks (e.g. width and depth) boosts the model performance if done correctly. However, given a neural network model, it is not straightforward to address the crucial question `how do we scale the network?'. In this talk, I will show how we can leverage different mathematical results to efficiently scale neural networks, with empirically confirmed benefits.

The conventional power system is transforming into a smart grid through the integration of information and communication technologies (ICT). The incorporation of advanced ICT in the smart grid enhances its control and operational efficiency.

AI has become a driving force for new scientific discovery. In this talk I will present our recent work in generative AI for sub-seasonal weather forecast, drug design and molecular modeling, where we outperformed state-of-the art prediction accuracy with dramatic reduction in computational resources. These works demonstrate the importance of the integration of AI with scientific problems and its transformative potential in both theoretical and practical applications.

Point-of-care (PoC) devices refer to portable and affordable diagnosing systems, mostly used in the field of healthcare. PoCs have garnered interest, particularly for detecting diseases that require multiple diagnostic tests and in resource-limited areas.

Time series clustering is an essential machine learning task with applications in many disciplines. While the majority of the methods focus on time series taking values on the real line, very few works consider time series defined on the unit circle, although the latter objects frequently arise in many applications. In this talk, the problem of clustering circular time series is discussed.

Until very recently, distribution-led local flexibility markets were exclusively an academic endeavor, with few practical applications, mostly limited to small-scale innovation projects. However, with European regulation finally catching up with the realities of modern distribution networks, local flexibility markets are slowly becoming a reality - new ones popping up across the continent, or some even becoming a BAU option in the most advanced countries.

The Pauli-Poisswell equation models fast-moving charges in semiclassical semi-relativistic quantum dynamics. It is at the center of a hierarchy of models from the Dirac-Maxwell equation to the Euler-Poisson equation that is linked by asymptotic analysis of small parameters such as the Planck constant or inverse speed of light. We discuss the models and their application in plasma and accelerator physics as well as the many mathematical problems they pose

This Dean's Distinguished Lecture is part of the ECE Graduate Seminar. Modern Artificial Intelligence (AI) technologies, led by deep learning, have gained unprecedented momentum over the past decade.

Brain activity over the entire network is complex. A full understanding of brain activity requires careful study of its multi-scale spatial-temporal organization. Motivated by these challenges, we will explore some characterizations of dependence between components of a multivariate time series and then apply these to the study of brain functional connectivity.

We consider the widely used continuous $Q_{k+1}-Q_k$ quadrilateral or hexahedral Taylor-Hood elements for the finite element discretization of the Stokes and generalized Stokes systems in two and three spatial dimensions.

Wave propagation and scattering problems are of huge importance in many applications in science and engineering - e.g., in seismic and medical imaging and more generally in acoustics and electromagnetics.

A subtle difference in a diffusion model can lead to an opposite conclusion. We need to understand how each component involved in the diffusion phenomenon contributes to the diffusion model. In this talk, we will discuss how nonconstant persistence and permeability play a role in the diffusion phenomenon.

Liquid metal(LM)-based electronics have the potential to shape the future of intelligent systems, soft robotics, and wearable technologies by leveraging their sensing, actuation, and computational capabilities. This talk will discuss methods to harness the unique properties of liquid metal for applications in wearable sensors, soft actuators, and reconfigurable electronic platforms.

The term doubly nonlinear refers to the fact that the diffusion part depends nonlinearly both on the gradient and the solution itself. Such equations describe several physical phenomena and were introduced by Lions and Kalashnikov. These equations have an intrinsic mathematical interest because they represent a natural bridge between the more natural generalizations of the heat equation: the p-Laplacian and the porous medium equation.

As many types of IoT devices worm their way into numerous settings and many aspects of our daily lives, awareness of their presence and functionality becomes a source of major concern. Hidden IoT devices can snoop (via sensing) on nearby unsuspecting users, and impact the environment where unaware users are present, via actuation.

In the first part of the talk, we introduce Ordered Dropout, a mechanism that achieves an ordered, nested representation of knowledge in deep neural networks (DNNs).

Recently, InGaN-based blue, green, and red (RGB) micro-LEDs have garnered significant attention and interest due to their exceptional features such as high contrast, intense brightness, excellent energy efficiency, and long device lifetimes, positioning them as strong contenders as the next-generation display technology.