DATA SCIENCEPhD Dissertation Proposal DefenseRuofan Hu Time: Sep 15, 2025, from 9:00- 10:00 amLocation: Gordon Library Conference Room 303PhD Committee:Prof. Elke A. Rundensteiner, Data Science, WPI. Advisor.Prof. Randy Paffenroth, Mathematical Sciences, WPI.Prof. Fabricio Murai, Data Science, WPI.Prof. Feifan Liu, Population and Quantitative Health Sciences, UMass Chan Medical School. External member. Title: Learning from Weak SupervisionAbstract: Deep learning models often rely on high-quality labeled data, yet such resources are scarce and costly in domains like public health and healthcare. This dissertation addresses the problem of learning from weak supervision, where labels are noisy, incomplete, or coarse. This dissertation focuses on two major regimes: noisy supervision, which involves developing strategies to learn effectively from mixed- and unknown-quality labels, and indirect supervision, which entails designing methods that leverage coarse-grained signals to guide fine-grained tasks, such as rationale extraction and clinical notes summarization. Across four tasks, novel approaches are proposed to reduce dependence on expert annotations, mitigate label noise and imbalance, and exploit existing high-level signals. This work contributes to the development of robust, scalable models that enhance clinical decision support and medical text understanding.

Seminar Series “Successes and Challenges in Using Mathematical Modeling to Bridge Scales in Muscle Contraction” Sam Walcott, PhD Mathematical Sciences Worcester Polytechnic Institute Monday, September 15, 2025 GP1002 12:00pm – 12:50pm Abstract: The contraction of muscle powers vital processes like locomotion, gastric motility, and blood circulation. The rational development of therapies for dysfunction in these processes, e.g. genetic heart disease, depends on connecting molecular-scale interactions to physiological function. Since the 1950s, the sliding filament and cross bridge theories seemed to relate molecular interactions to the contraction of muscle cells. But, despite the successes of these theories, aspects of muscle contraction, e.g. the history dependence of isometric force, have eluded a mechanistic description. I'll discuss our work using mathematical modeling to bridge the molecular to cellular and larger scales. In particular, I'll share three stories: first, how we used mathematical modeling to explain how a molecular change causes fruit flies to jump further. Second, how our model of calcium activation was used to understand the mechanism of a heart drug. And third, how our modeling suggests a molecular mechanism for muscle's history dependence, and how that ties into a newly-discovered mechanism of muscle activation. Biography: (last name pronounced like the words "wall" and "cot"). I received my undergraduate degree in Biology and my PhD in Theoretical and Applied Mechanics, both from Cornell University. I did my first postdoc with David Warshaw at the University of Vermont, performing measurements in the laser trap to understand smooth muscle regulation. I did my second postdoc with Sean Sun at Johns Hopkins, developing mathematical models for cell mechanosensation. I then became an independent investigator at UC Davis in the Department of Mathematics. I am currently at Worcester Polytechnic Institute in the Department of Mathematical Sciences. My primary research interest is using mathematical modeling to tie together experimental measurements of biological systems at the molecular, cellular, and larger scales. For a zoom link please contact Kate Harrison at kharrison@wpi.edu

Ludwig Boltzmann was unsatisfied with fluid theory (Navier-Stokes Eq., etc.) as it lacked a microscopic theory that could explain the irreversibility at the core of thermodynamics. He committed himself to show how the disbelieved notion of atoms could unravel fluid dynamics into a stochastic particle theory. This talk is about my research program to similarly find a microscopic and ballistic theory of “particles” that generates quantum mechanics, while staying within the guard rails of Einstein’s notion of “realism”. We will visit several less traveled areas of quantum theory and mathematics while following a path to this goal and end with a demonstration of a computational demonstration.David Cyganski Professor Emeritus, Electrical and Computer Engineering, RBE - WPIAll Faculty and students are welcome and encouraged to join us! Physics Department Room 218

Life Sciences Seminar Series "Investigating the post-transcriptional role of regulatory small RNAs in mycobacterial stress responses" Opeyemi Isaac Ibitoye, BBT PhD Candidate, Advisor: Associate Professor Scarlet Shell"Identification of Functional Targets Reveals Novel GLD-2 Functions in the C. elegans Proximal Germline" Danni Li, BBT PhD Student, Advisor: Karl-Frédéric Vieux Tuesday September 16th 2025 @ 12:00 PM 60 Prescott St. Gateway Park 1st floor GP1002

Live Stats

Department of Mathematical SciencesDiscrete Math SeminarTuesday, September 16th, 20254:00PM-4:50PMOlin Hall 126Speaker: Jane Coons, WPITitle: Quasi-Independence Models with Rational Maximum Likelihood Estimator Abstract: Let X and Y be random variables. Quasi-independence models are log-linear models that describe a situation in which some states of X and Y cannot occur together, but X and Y are otherwise independent. We characterize which quasi-independence models have rational maximum likelihood estimator, or MLE, based on combinatorial features of the bipartite graph associated to the model. In this case, we give an explicit formula for the maximum likelihood estimate. We also give a substantial introduction to algebraic statistics, which will be accessible to advanced undergraduate students.