The Computer Science (CS) department welcomes prospective students interested in our graduate programs to our annual Prospective Student Visit Day on Friday, Nov. 7. We are looking for strong students with diverse backgrounds to join our MCS and PhD programs. A wide variety of research areas is represented by our world-class faculty including algorithms, computational epidemiology, distributed computing, human-computer interaction, machine learning, massive data algorithms and technology, mobile computing, networks, programming languages, text mining, security, and virtual reality. Write to our Graduate Program Administrator at cs-grad@uiowa.edu if you'd like to visit.
An important part of the prospective Student Visit Day is the 11th Iowa Computer Science Graduate Research Symposium (2025), in which senior PhD students will present talks showcasing their current research. This will be followed by a keynote by a CS faculty member. Talks are intended for a wide audience with interest in CS, including CS juniors and seniors. The talks presented by current CS graduate students at the Symposium are excellent examples of the exciting CS research taking place here on the UI campus!
-------------------
Schedule
All times CT [Conversion to your timezone possible here if needed]
Friday, Nov. 7, 2025
Morning Session (By invitation only)
Overview of Graduate Programs by Professor and Director of Graduate Studies Steve Goddard
Short Faculty Research Presentations by University of Iowa CS Professors Katherine Kosaian, Naimul Hoque, Taylor Olson, and Weiran Wang.
_________________________
Afternoon Sessions (Everyone invited for afternoon sessions! Graduate students encouraged to attend.)
Graduate Student Research Symposium Sessions / PhD Student Presentations
1:30-3:30 p.m.
_________________
1:30 p.m. Apoorv Ingle, PhD Student - Overloading Reloaded: Making Ad-hoc Polymorphism Even Less Ad-hoc
2 p.m. Chen Sun, PhD Student - Closing Regulatory Gap: Reducing Information Fragmentation and Enabling Scalable Enforcement
2:30 p.m. Joshua Sobel, PhD Student - Distributed Sampling and Counting Algorithms
3 p.m. Jamil Gafur, PhD Student - What happens when you lobotomize a neural network?
___________________
3:30-4 p.m. - Break/reception
________________
4-5 p.m. Keynote Speaker: Assistant Professor Mehrdad Moharrami
Topic: Recovering Planted Structures in Randomly Weighted Graphs
---------------------------------------
More on Our Speakers:
Title: Overloading Reloaded: Making Ad-hoc Polymorphism Even Less Ad-hoc
Abstract: Programmers take overloading in modern programming languages for granted. In this talk I will first compare the current mechanisms of overloading implemented in modern day compilers. I will then describe how these mechanisms fail with certain declarative features (multiparameter type classes and type families) in an advanced declarative functional programming language (GHC/Haskell). Finally, I will sketch a design of a new core language (System FD) that promises to resolve these issues.
6th-year PhD student | Advisor: J. Garrett Morris | Areas of research: Functional Programming Languages, Compilers
_______________
Title: Closing Regulatory Gap: Reducing Information Fragmentation and Enabling Scalable Enforcement
Abstract: Data-driven platforms, ad-tech, and creator economies increasingly mediate how personal data is collected, monetized, and governed. While laws such as the GDPR and CCPA, and agencies like the FTC, work to protect consumers, enforcement struggles to keep pace, creating a persistent gap between legal obligations and real-world practice. My research traces this gap to two key challenges: information fragmentation and the need for enforcement at scale. In the first part of the talk, I present the challenge of information fragmentation, where crucial knowledge about data practices is siloed within firm units, and the regulatory precedents needed for compliance are scattered across dozens of authorities. To solve this, we propose mandatory, signed disclosures to force internal clarity for firms and create an automated archive to centralize external regulatory knowledge. In the second part of talks, we move forward to enable enforcement at scale. I demonstrate an automatic tool for detecting affiliate link and its disclosure, enabling large-scale auditing of the influencer economy.
5th-year PhD student | Advisor: Rishab Nithyanand | Areas of research: Online Privacy
_______________
Title: Distributed Sampling and Counting Algorithms
Abstract: A wide variety of algorithms have emerged for sampling and counting combinatorial structures defined on graphs, for example: spanning trees, colorings, and independent sets. The most important and fundamental technique is MCMC (Markov chain Monte Carlo). I will discuss my work in bringing these same techniques to distributed computing models. In particular, I will highlight a new distributed algorithm for taking fast random walks and sampling random spanning trees. I will also discuss new connections between Markov chain sampling and matrix multiplication.
5th-year PhD student | Advisor: Sriram Pemmaraju | Areas of research: Theoretical computer science, Markov chains, distributed computing
_______________
Title: What happens when you lobotomize a neural network?
Abstract: Over the course of a few years, neural networks have grown in size, causing a strain on computational and memory resources. Iterative Magnitude Pruning (IMP) is a simple and widely used method to reduce model size by removing low-magnitude weights while preserving accuracy, but the mechanisms behind its effectiveness remain underexplored. In our work, we analyze IMP across ResNet20, VGG16, and RegNetX on CIFAR10 and TinyImageNet, finding that it preferentially prunes deeper layers while preserving early feature extractors, and often leaves behind a small number of weights per neuron even at high sparsity's. As sparsity increases, the network becomes more sensitive to further pruning, indicating growing fragility. We attempt to take advantage of this by implementing network layer fusion which collapses the highly sparse layers into a smaller dense layer, thus offering better resource efficiency.
5th-year PhD student | Advisor: Steve Goddard | Areas of research: AI/ML, Energy Efficient Machine Learning
_______________
Keynote Speaker: Mehrdad Moharrami
Title: Recovering Planted Structures in Randomly Weighted Graphs
Abstract: We study the recovery of hidden structures in randomly weighted graphs, focusing on two canonical cases: the planted matching and the planted spanning tree. In the matching problem, a perfect pairing of vertices is hidden among random edges; we show that when the signal is strong, near-perfect recovery is possible, while weaker signals lead to a sharp threshold where only partial recovery can be achieved. In the spanning tree problem, a hidden spanning tree is embedded in a randomly weighted complete graph. We analyze the performance of the minimum spanning tree algorithm, characterize the fraction of the tree that can be recovered, and extend Frieze's zeta(3) result to this planted setting. Across both problems, our results identify the boundaries between exact recovery, partial recovery, and failure, establishing when efficient algorithms succeed through techniques from probability, graph theory, and combinatorial optimization.
Bio: Mehrdad Moharrami is an Assistant Professor in Computer Science at the University of Iowa. Previously, he was a TRIPODS Postdoctoral Research Fellow at UIUC, working with Prof. R. Srikant. He earned his BSc in Mathematics and Electrical Engineering from the Sharif University of Technology, Iran, and holds MSc degrees in Electrical Engineering and Mathematics from the University of Michigan. In Winter 2020, he received his PhD in Electrical Engineering, earning the Rackham Predoctoral Fellowship for his dissertation.