Speaker: Thu Nguyen, Department of Math and Statistics, University of Maryland, Baltimore County

Time: Wednesday, April 1, from 2:30 PM to 3:30 PM

Location for in-person participants: 1146 FAB 

Zoom link for online audience: https://wayne-edu.zoom.us/j/92845590121?pwd=CpRA5Wa5gzSMn2xiVkR2abD83O5nrH.1

Title: Bridging the Gap: From Classical Time Series to Transformers Based Models

Abstract: Time series forecasting increasingly requires balancing the interpretability of classical models with the representational power of deep learning. While traditional “white box” methodologies like ARIMA offer structural transparency, they often struggle with the complex, non-linear dynamics of high-dimensional datasets. Conversely, standard Transformers are frequently hindered by a “quadratic trap”—an O(L^2) complexity that makes long-sequence forecasting computationally prohibitive. This talk introduces a class of Transformer-based models designed to close this analytical gap by incorporating inductive biases from classical theory, such as frequency-domain analysis and exponential smoothing. We present ProbSparse Self-Attention, which utilizes KL-divergence to reduce complexity to O(L log L), and a segment-based approach that shifts from isolated points to local neighborhood similarity. Furthermore, we utilize a Generative Style Decoder for single-forward-pass inference to drastically accelerate long-horizon forecasting. Empirical evaluations on ETT and Exchange Rate benchmarks demonstrate significant improvements over conventional models like Prophet and DeepAR. 

Bio: Dr. Thu Nguyen is an Assistant Professor in the Department of Math and Statistics at the University of Maryland, Baltimore County (UMBC). She obtained her PhD degree in Statistical Signal Processing from The Lille 1 University of Science and Technology in 2014, and her PhD degree in Applied Mathematics from Wayne State University in 2020.  Dr. Nguyen's research interests lie at the intersection of stochastic approximation, Monte Carlo methods, stochastic systems, applied mathematics, and their applications. Her current research focuses on designing and analyzing new stochastic approximation algorithms for stochastic networked systems, as well as developing and applying efficient Monte Carlo samplers for complex statistical problems. She is also interested in applying these techniques to problems in statistical learning, artificial intelligence, and machine learning. Since joining UMBC in 2020, Dr. Nguyen has expanded her research interests in the fields of machine learning and deep learning. Her research in this area revolves around the development of a novel class of deep learning-based models tailored to diverse time series modeling tasks.

______________________________________________________________________

AIDaS: CAD Seminar Series

Advancing Knowledge, Innovation, and Collaboration in Computation, AI, and Data Science (CAD)

The CAD Seminar Series is a primary seminar series at Wayne State University’s Institute for AI and Data Science (AIDaS). It is a dedicated platform for advancing knowledge, fostering innovation, and promoting collaboration across the fields of Computation, Artificial Intelligence, and Data Science. This series brings together leading experts, researchers, and professionals to explore the latest developments, tackle emerging challenges, and drive forward-thinking solutions at the convergence of these critical disciplines.

Objectives:

• Advance Knowledge: Share cutting-edge research and insights that push the boundaries of what is known in CAD.

• Foster Innovation: Encourage the development of novel ideas and solutions through interdisciplinary dialogue and creative thinking.

• Promote Collaboration: Unite expertise across disciplines and build bridges between academia, industry, and government to address complex problems and create opportunities for joint ventures.

Target Audience: The CAD Seminar Series is designed for a diverse audience, including faculty, researchers, students, and professionals in Computation, AI, Data Science, and related fields. It serves as a forum for exchanging ideas, networking, and contributing to the growth of these rapidly evolving areas. We highly recommend in-person attendance to enhance engagement and networking opportunities with speakers and fellow participants.

Call for Participation: We welcome contributions from researchers, practitioners, and students. Whether presenting your work, participating in discussions, or attending as a learner, your involvement is crucial to the success of this collaborative initiative.