Select a calendar:
Filter April Events by Event Type:
University Calendar
Events for April
-
PhD Thesis Proposal - Grace Zhang
Tue, Apr 01, 2025 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Knowledge Transfer for Data Efficient Reinforcement Learning
Committee : Gaurav Sukatme (Chair), Stefanos Nikolaidis, Erdem Biyik, Daniel Seita, Stephen Tu
Abstract: Reinforcement learning and the closely related inverse reinforcement learning problems are general and powerful frameworks to learn sequential decision making tasks with only a reward function or demonstrations and minimal assumptions on the environment. However, the trade-off is that these algorithms can be very data inefficient, in the number of trials required in the training environment or the number of demonstrations required. In my work I explore how to achieve more data efficient learning through knowledge transfer between environments or between tasks. Specifically, on how to transfer behaviors between environments, how to share behaviors between tasks in multi-task RL, and how to utilize multi-task information to do inverse RL from limited demonstrations.
Location: Ginsburg Hall (GCS) - 402C - 4th Floor
Audiences: Everyone Is Invited
Contact: Ellecia Williams
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Dissertation Defense - Hanchen Xie
Wed, Apr 02, 2025 @ 12:00 PM - 02:00 PM
Thomas Lord Department of Computer Science
University Calendar
Dissertation Title: Mitigating Environment Misalignment And Discovering Intrinsic Relations Via Symbolic Alignment
Date and Time: April 2, 12 pm to 2 pm.
Location: RTH 217
Committee: Yue Wang (Chair), Wael Abd-Almageed, Aram Galstyan, Emilio Ferrara, Peter Beerel
Abstract: Deep learning models have achieved remarkable success on various computer vision tasks. Modern state-of-the-art methods can not only recognize the visual appearance of objects but also discover intrinsic relations of objects (e.g., dynamics or causal relations). However, collecting sufficient training data for the intrinsic relations can be expensive or infeasible in many scenarios, such as car incident videos in the real world. As an alternative, one can generate data in a different environment, such as synthetic data, that depicts the same intrinsic relations. Yet, end-to-end models may suffer from environment misalignment challenges, such as visual domain or environment context shift, so the model generality is limited. To mitigate such misalignment challenges, we propose symbolic alignment, a novel learning strategy that utilizes a common symbolic space to align various environments. We first conduct a case study on dynamics prediction to reveal the environment misalignment challenges on our proposed datasets. Next, to obtain insight into the challenge, we provide an investigation of the implicit position encoding in the dynamics prediction model. Then, we present a learning framework that separates the learning of appearance recognition and dynamics relations discovery to improve the generality of the dynamics prediction model. Then, we generalize the symbolic alignment strategy and introduce a novel framework, Look, Learn, and Leverage L3. L3 decomposes the learning process into three distinct phases and achieves promising results on three intrinsic relations discovery tasks. Finally, we extend the environment misalignment discussion to video classification and demonstrate the potential of symbolic alignment to mitigate the video content inconsistency between training and inference.Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Hanchen Xie
Event Link: https://usc.zoom.us/j/95803531086?pwd=LwnaIMsMv44jIIkvvlUEXD3gAbqb2N.1
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Soumyaroop Nandi
Fri, Apr 04, 2025 @ 02:45 PM - 04:45 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Context-Aware Semantic Forgery Detection in Biomedical and Natural Images
Committee: Dr. Premkumar Natarajan (Chair), Dr. Emilio Ferrara, Dr. Daniel O’Leary, Dr. Erdem Biyik, and Dr. Gale Lucas
Abstract:Copy-move forgery is one of the most common and challenging forms of image manipulation, where regions within an image are duplicated and repositioned to conceal or falsify visual evidence. Detecting these manipulations becomes especially difficult in the case of semantic or context-aware forgeries, where duplicated content is strategically placed to mislead interpretation or alter meaning. This challenge is further compounded in specialized domains such as biomedical imaging, where image tampering can undermine scientific integrity by distorting experimental results. In the proposed thesis, we explore and develop state space model-based attention networks to advance the detection of copy-move and semantic image forgeries in both natural and biomedical images. We begin by introducing a visual state space modeling approach that uses normalized attention maps to locate and compare similar regions within an image. A region-based block-attention mechanism, integrated with this model, enables precise identification of manipulated and authentic areas, producing detailed localization maps of both the source and duplicated regions. To address the limitations of existing datasets, we propose a comprehensive copy-move forgery detection dataset designed to capture a wider range of sophisticated tampering techniques. Furthermore, we extend our methods to biomedical images, leveraging state space models as similarity detectors that focus on duplicated regions, enabling effective detection of manipulations that traditional models often fail to identify. This thesis aims to advance the field of semantic forgery detection by providing efficient and robust techniques for identifying both low-level pixel alterations and high-level, context-driven forgeries across diverse imaging applications.
Location: Ginsburg Hall (GCS) - 402C - 4th Floor
Audiences: Everyone Is Invited
Contact: Ellecia Williams
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
"Keys to Life" series at USC ORSLMon, Apr 07, 2025 @ 12:00 PM - 01:00 PM
USC Viterbi School of Engineering
University Calendar
"Keys to Life" with Prof. Weiss is a motivational discussion series designed to promote student success and well-being. This series is for students who want to develop their "keys" in a small group setting and a peaceful, reflective environment. Finding purpose is essential to living a meaningful life and key to personal fulfillment. This series will help students identify and articulate their purpose and provide group motivation to work towards it. A unique feature of the series will be its peripatetic "Purpose Walks" through campus.
More Information: Keys to Life with Prof. Weiss.jpg
Location: University Religious Center (URC) - courtyard
Audiences: Everyone Is Invited
Contact: Elisabeth Arnold Weiss
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Arash Hajisafi
Wed, Apr 09, 2025 @ 12:30 PM - 02:00 PM
Thomas Lord Department of Computer Science
University Calendar
Presentation Title: Dynamic GNNs for Accurate and Efficient Modeling of Instant and Lagged Dependencies in Multivariate Time Series
Date and Time: Wednesday, April 9th, 2025 - 12:30p - 2:00p
Location: GCS 502C
Committee Members: Cyrus Shahabi (Chair), Ibrahim Sabek, Viktor Prasanna, Ruishan Liu, John P. Wilson (External)
Abstract: Graph Neural Networks (GNNs) have shown great success in modeling complex dependencies within multivariate time series by explicitly capturing intra-series (within individual series) and inter-series (across different series) relationships. However, existing methods often struggle to represent evolving correlations, particularly when multiple contexts and lagged interactions are involved. My previous research has developed GNN-based prediction models addressing instant dependencies across various contexts, incorporating both static and dynamic relationship aspects, and achieving significant improvements in forecasting accuracy and efficiency. Despite these advancements, real-world time series, such as those found in financial markets, frequently exhibit lagged dependencies, where changes in one series influence others after varying delays. Building on my prior contributions, my dissertation proposes developing a novel dynamic GNN method explicitly designed to capture these lagged dependencies, aiming to further enhance the prediction accuracy in applications like stock forecasting.Location: Ginsburg Hall (GCS) - 502C
Audiences: Everyone Is Invited
Contact: Arash Hajisafi
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Jiahao Wen
Thu, Apr 10, 2025 @ 12:00 PM - 01:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title of Thesis Proposal: Optimal r-Adaptive In-Timestep Remeshing for Elastodynamics
Date and Time: April 10th, 12 pm - 1pm
Location: SAL 213
Committee Members: Prof. Jernej Barbic, Prof. Yong Chen, Prof. Oded Stein, Prof. Satyandra Gupta, and Prof. Stefanos Nikolaidis.
Abstract: This work is about finding optimal degrees of freedom for FEM simulation of nonlinear deformable objects with frictional contacts. This is done by moving the vertices in the undeformed (reference) mesh to improve the match to the true analytical solution of the underlying PDE. I.e., get closer to the true solution with a fewer number of mesh vertices by optimally repositioning those vertices in the undeformed mesh. More broadly, the work tries to improve how partial differential equations are solved by adapting the FEM solution space.Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: Jiahao Wen
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Robby Costales
Thu, Apr 10, 2025 @ 03:00 PM - 04:30 PM
Thomas Lord Department of Computer Science
University Calendar
Title: The Three-Tiered Exploration Problem in Open-Ended Adaptive Learning
Committee members: Stefanos Nikolaidis (chair), Erdem Biyik, Stephen Tu, Willie Neiswanger, Daniel Seita
Abstract: A central challenge in training adaptive decision-making agents via meta-reinforcement learning (meta-RL) is meta-exploration—the search for an efficient exploration strategy that generalize to new, unseen tasks. Another bottleneck is the significant expense of manually designing training task distributions. While autocurricula methods—which automatically generate appropriately challenging training tasks for the learning agent—are well-studied in the standard RL setting, their application to meta-RL has been underexplored. These autocurricula approaches are a promising route for both (1) reducing the difficulty of meta-exploration and (2) removing the need for hand-designing tasks for meta-RL training, but the emergent training dynamics are complex—with each component mutually exacerbating each others' separate instabilities. In this talk, I outline a preliminary framework for understanding this combined learning problem, and present a research trajectory for addressing the associated challenges, building on my ongoing PhD work.Location: Ginsburg Hall (GCS) - 502C - 5th Floor
Audiences: Everyone Is Invited
Contact: Ellecia Williams
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
DREAM Industry Mentorship speaker series- special event with Teague Egan
Mon, Apr 14, 2025 @ 10:00 AM - 11:00 AM
USC Viterbi School of Engineering
University Calendar
DREAM connects students with experienced industry professionals from a variety of tech and destination companies who help them create a vision for their futures, align their careers around purpose, and build character in the context of growth, reinvention, and constant change. Industry mentors discuss how professional challenges present opportunities for character and leadership development. This event features visionary USC alum Teague Egan, the Founder and CEO of EnergyX, discussing his remarkable career as an entrepreneur and energy futurist developing cutting-edge lithium and battery technology.https://eis.usc.edu/dream/
More Information: DREAM Flyer 4-14 Teague Egan talk.png
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Elisabeth Arnold Weiss
Event Link: https://cglink.me/2nB/r403917
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
"Keys to Life" series at USC ORSL
Mon, Apr 14, 2025 @ 12:00 PM - 01:00 PM
USC Viterbi School of Engineering
University Calendar
"Keys to Life" with Prof. Weiss is a motivational discussion series designed to promote student success and well-being. This series is for students who want to develop their "keys" in a small group setting and a peaceful, reflective environment. Finding purpose is essential to living a meaningful life and key to personal fulfillment. This series will help students identify and articulate their purpose and provide group motivation to work towards it. A unique feature of the series will be its peripatetic "Purpose Walks" through campus.
More Information: Keys to Life with Prof. Weiss.jpg
Location: University Religious Center (URC) - courtyard
Audiences: Everyone Is Invited
Contact: Elisabeth Arnold Weiss
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Changzhi Xie
Mon, Apr 14, 2025 @ 02:30 PM - 03:30 PM
Thomas Lord Department of Computer Science
University Calendar
Title of Presentation: On the Dynamics of Learning Linear Functinos with Neural Networks
Date and Time: 4.14 2:30-3:30PM
Location: EEB 203
Committee Members: Mahdi Soltanolkotabi(committee chair), Haipeng Luo, Robin Jia, Vatsal Sharan, Adel Javanmard.
Abstract: We study the gradient descent training dynamics of fitting a one-hidden-layer network with multi-dimensional outputs to linear target functions. That is, we focus on a realizable model where the inputs are drawn i.i.d. from a Gaussian distribution and the labels are generated according to a planted linear model with multiple outputs. This framework serves as a good model for a variety of interesting problems including end-to-end training in inverse problems and various auto-encoder models in machine learning. Despite the seemingly simple formulation, understanding training dynamics is a challenging unresolved problem. This is in part due to the fact that the training landscape contains multiple local optima and it is completely unclear why gradient descent from random initialization is able to escape such bad optima. In this work, we develop the first comprehensive analysis of the gradient descent dynamics for learning linear target functions with ReLU networks. We show that gradient descent with moderately small random initialization converges to a global minimizer at a linear rate. To rigorously show that GD avoids local optima, we develop intricate techniques to decompose the loss and control the GD trajectory, which may have broader implications for the analysis of non-convex optimization problems involving local optima. We corroborate our theoretical results with extensive experiments with various configurations.Location: Hughes Aircraft Electrical Engineering Center (EEB) - 203
Audiences: Everyone Is Invited
Contact: Changzhi Xie
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Tejas Srinivasan
Mon, Apr 14, 2025 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title of Thesis Proposal: Facilitating Reliable Human-AI Collaboration Under Uncertainty
Date and Time: April 14, 2025, 4--5pm
Location: GCS 402C
Committee Members: Jesse Thomason (Chair), Robin Jia, Heather Culbertson, Morteza Dehghani, Diyi Yang
Abstract: AI systems are increasingly assisting humans with decision-making tasks. Effective human-AI collaboration requires AI assistants to be reliable by not only being accurate but also knowing when they don’t know and acting appropriately when uncertain. Popular strategies for handling uncertainty include abstaining from answering, providing prediction sets using conformal prediction, communicating uncertaintyto users, and asking clarification questions to resolve uncertainty. However, these mechanisms do not always facilitate appropriate reliance on and utilization of AI systems by users. In this thesis, we explore methods for proactively mitigating under- and over-reliance in human-AI collaboration under uncertainty. In selective prediction, always abstaining when uncertain can lead to under-utilization by the user, so we develop an algorithm to reduce over-abstention in multimodal selective prediction systems without increasing the error rate of the system’s predictions. When communicating uncertainty, we find that user trust can bias how users rely on AI confidence estimates and lead to inappropriate reliance, which we mitigate by adapting AI assistants’ behavior to user trust levels. Finally, we propose reducing over-reliance on LLM agents by modeling and proactively resolving uncertainty about user goals through frictive dialogue. Our works highlight the importance of modeling uncertainty about AI predictions and the user-AI interaction itself, and the benefits of responding to uncertainty through AI introspection and adaptive AI behaviorsLocation: Ginsburg Hall (GCS) - 402C
Audiences: Everyone Is Invited
Contact: Tejas Srinivasan
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Sara Babakniya
Tue, Apr 15, 2025 @ 12:00 PM - 01:30 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Foundational Studies of Privacy and Efficiency in Federated Machine Learning
Date and Time: Tuesday, April 15th, 2025 - 12:00p - 1:30p
Location: EEB539
Committee Members: Prof. Salman Avestimehr (Chair), Prof. Harsha V. Madhyastha, Prof. Jose-Luis Ambite, Prof. Sai Praneeth Karimireddy, Prof. Mahdi Soltanolkotabi
Abstract: Federated learning (FL) enables collaborative model training across distributed devices while preserving user data privacy. However, deploying FL in practice has challenges, such as limited client resources, communication overhead, privacy concerns, and data heterogeneity.
My research addresses these fundamental barriers by developing general and adaptable frameworks that make FL more efficient and scalable in real-world environments. First, I discuss catastrophic forgetting in federated class-incremental learning, where a client's local data distribution may shift over time. I propose a data-free generative replay framework that does not require extra data storage or sharing from the clients. Then, I present my work that explores how to reduce the communication and computation costs of federated training while preserving model performance. I show that we can incorporate sparse learning to reduce costs, but we must carefully coordinate the local and global sparsity patterns.
Building on my prior knowledge of privacy and efficiency, I propose an efficient method to fine-tune language models on edge data. The state-of-the-art language models have been trained on the majority of the available public data. Therefore, these models need to be trained on the users' private data to improve their performance further. I investigate how we can move towards this goal without compromising privacy or efficiency.Location: Hughes Aircraft Electrical Engineering Center (EEB) - 539
Audiences: Everyone Is Invited
Contact: Sara Babakniya
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
DREAM Industry Mentorship speaker series- with Binti Yost
Wed, Apr 16, 2025 @ 09:00 AM - 10:00 AM
USC Viterbi School of Engineering
University Calendar
DREAM connects students with experienced industry professionals from a variety of tech and destination companies who help them create a vision for their futures, align their careers around purpose, and build character in the context of growth, reinvention, and constant change. Industry mentors discuss how professional challenges present opportunities for character and leadership development. This event features Binti Yost, Principal at KPMG- Economic and Valuation Services, sharing insights from her career in consulting for Fortune 500 companies. https://eis.usc.edu/dream/
More Information: DREAM Flyer 4-16 Binti Yost.png
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Elisabeth Arnold Weiss
Event Link: https://cglink.me/2nB/r403700
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Mi-Ying Miryam Huang
Thu, Apr 17, 2025 @ 12:00 PM - 01:30 PM
Thomas Lord Department of Computer Science
University Calendar
Presentation Title: Towards Publicly Verifiable Cryptography: Obfuscation, Fully Homomorphic Encryption, and Proof Carrying State.
Date and Time: April 17th 12:00pm to 1:30pm
Location: Ginsburg 503C
Committee members: David Kempe, Greta Panova (math department), Vatsal Sharan, Shanghua Teng, Jiapeng Zhang
Abstract: We explore public verifiability in cryptography. This proposal highlights two main results and one ongoing research direction:
Through a quantum lens, we introduce Quantum Obfuscation for approximate Unitary Quantum Functionality. By using advanced quantum techniques, our construction supports approximate unitary quantum functionalities with quantum inputs and outputs, significantly extending beyond existing limitations by Bartusek et al (STOC 2023, STOC 2024). Utilizing Quantum Teleportation combined with Projective Linear Measurement (PLM) quantum programs, we overcome critical obstacles from previous works and open potential applications in quantum copy-protection, quantum functional encryption, and secure quantum software distribution.
From a classical cryptographic perspective, we develop a Publicly Verifiable Fully Homomorphic Encryption (pvFHE) scheme, building upon the FHEW framework by Ducas and Micciancio (Eurocrypt 15). Integrating the GINX homomorphic accumulator, our scheme improves efficiency during bootstrapping and verification. Moreover, we introduce a generalized Rank-1 Constraint System (Ring R1CS) and construct a succinct non-interactive argument (SNARG). This approach provides efficient verifiability and strong security guarantees, including enhanced client data privacy, adhering to the recently introduced privacy framework by Cini et al. (Crypto 24).
Finally, our ongoing project, Proof-Carrying Quantum States, further extends these concepts to achieve verifiable quantum computations, bridging classical and quantum cryptographic techniques to ensure computation integrity and privacy. Together, these contributions advance both theoretical foundations and practical applications of publicly verifiable cryptographic protocols.Location: Ginsburg Hall (GCS) - 503C
Audiences: Everyone Is Invited
Contact: Mi-Ying Miryam Huang
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Rajrup Ghosh
Mon, Apr 21, 2025 @ 11:30 AM - 01:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title of Thesis Proposal: Enabling Volumetric Video Conferencing and Live Streaming
Date and Time: April 21, 2025 (Monday), 11:30 am - 1 pm PST
Location: 302C, GCS, Floor 3
Committee Members: Ramesh Govindan (Chair), Barath Raghavan, Yue Wang, Harsha V. Madhyastha, Antonio Ortega
Abstract: Volumetric video streaming represents the next frontier in media communication, enabling highly immersive experiences by capturing and transmitting dynamic 3D scenes in real-time. However, significant challenges remain before widespread adoption is possible, including managing substantial bandwidth demands, achieving low-latency, supporting multi-party interactions, and maintaining high visual realism. This thesis explores novel methods to overcome these barriers. To enable immersive two-party conferencing, I introduce LiVo, a system that efficiently streams full-scene volumetric videos by extending 2D video encoding techniques, adaptive bandwidth allocation, and real-time view prediction and culling. LiVo significantly reduces bandwidth usage while maintaining an end-to-end latency of approximately 250 ms at 30 frames per second. Extending these capabilities to multi-party scenarios, I propose LiVo++, which must address challenges in synchronization, computation overhead, and changing network conditions. It proposes dynamic strategies to adapt video quality levels according to participants' device capabilities and available bandwidth, enabling robust and scalable interactive experiences. Finally, I propose to improve visual quality through LiVoGS, a system leveraging Gaussian Splatting. LiVoGS will explore integrating Gaussian Splatting with motion-compensated encoding inspired by traditional 2D codecs, to achieve photorealistic visual quality and significantly improved bandwidth efficiency. Collectively, these innovations enable practical, efficient, and visually realistic volumetric video conferencing and streaming, paving the way toward the future of immersive multimedia communication.Location: Ginsburg Hall (GCS) - 302C
Audiences: Everyone Is Invited
Contact: Rajrup Ghosh
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
"Keys to Life" series at USC ORSL
Mon, Apr 21, 2025 @ 12:00 PM - 01:00 PM
USC Viterbi School of Engineering
University Calendar
"Keys to Life" with Prof. Weiss is a motivational discussion series designed to promote student success and well-being. This series is for students who want to develop their "keys" in a small group setting and a peaceful, reflective environment. Finding purpose is essential to living a meaningful life and key to personal fulfillment. This series will help students identify and articulate their purpose and provide group motivation to work towards it. A unique feature of the series will be its peripatetic "Purpose Walks" through campus.
More Information: Keys to Life with Prof. Weiss.jpg
Location: University Religious Center (URC) - courtyard
Audiences: Everyone Is Invited
Contact: Elisabeth Arnold Weiss
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
DREAM Industry Mentorship speaker series- with Mehrad Noori
Wed, Apr 23, 2025 @ 11:00 AM - 12:00 PM
USC Viterbi School of Engineering
University Calendar
DREAM connects students with experienced industry professionals from a variety of tech and destination companies who help them create a vision for their futures, align their careers around purpose, and build character in the context of growth, reinvention, and constant change. Industry mentors discuss how professional challenges present opportunities for character and leadership development. This event features Mehrad Noori, Executive Producer at Reality Labs at Meta, sharing insights from his journey from undergraduate at USC School of Cinematic Arts and M.A. at Iovine and Young Academy to leading immersive content development as Executive Producer at NBC Universal, AnythingEverything, and Meta. https://eis.usc.edu/dream/
More Information: DREAM flyer 4-23 Mehrad Noori.png
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Elisabeth Arnold Weiss
Event Link: https://cglink.me/2nB/r403861
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Neel Patel
Thu, Apr 24, 2025 @ 12:30 PM - 01:30 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Combinatorial Optimization under Uncertainty, Incentives and Correlations.
Date and Time: 04/24, 12:30-1:30 pm
Location: GCS 502C
Committee Members: Shaddin Dughmi, David Kempe, Vatsal Sharan, Evi Micha and Greta Panova
Abstract:
This proposal considers algorithms for combinatorial problems, primarily those concerned with combinatorial selection under uncertainty and incentives, also known as stochastic selection problems. The core focus is on the two pivotal stochastic selection problems that include contention resolution schemes (CRS) and generalized prophet inequalities. Our contributions are twofold:
Our first contribution deepens the understanding of the stochastic selection problems beyond independent priors on the input and its implications on the famous matroid secretary conjecture. Our results completely characterize the CRS and prophet inequalities on matroids for pairwise independent priors. En route to proving our results, we develop techniques to sample exact pairwise independent vectors over a finite field from approximate pairwise independent vectors which later becomes a key ingredient for characterizing the difficult instance for binary matroid secretary conjecture.
The rest of the proposal aims to push the applications of the powerful algorithmic toolkit --- stochastic selection with a broader goal of identifying the algorithmic and economic questions that appear to be complex and algorithmically challenging, for which the techniques developed by online stochastic selection provide an alternative outlook, leading to more efficient and powerful algorithmic results. In this context, we prove the following key results:
1.) We obtain the first combinatorial generalized stationary prophet inequalities where our main result shows that the (offline) CRS plays a central role in the (online) stationary prophet inequality problem. This intriguing connection allows us to obtain several new algorithmic results as well as improves the existing results significantly.
2.) We systematically generalize the sparsification of stochastic matching problems to the general combinatorial structure. Here, we show that any combinatorial structure that exhibits `good’ CRS also exhibits strong stochastic sparsifiers.
3.) We obtain constant approximate delegation mechanisms for the principal-agent delegation problem with probing cost for a large class of combinatorial constraints. We obtain these mechanisms by reducing the delegation problem to the online version of CRS for combinatorial constraints.Location: Ginsburg Hall (GCS) - 502C
Audiences: Everyone Is Invited
Contact: Neel Patel
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - Zhuojin Li
Fri, Apr 25, 2025 @ 02:00 PM - 04:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Performance Modeling and Optimization for Machine Learning Systems: from Cloud Training to Edge Inference
Date and Time: Fri, April 25, 2-4pm
Location: EEB 403
Committee Members: Leana Golubchik (Chair), Murali Annavaram, Peter Beerel, Jyotirmoy V. Deshmukh, William G. J. Halfond
Abstract: Deep neural networks (DNNs) have achieved remarkable success in a wide range of tasks, from computer vision to natural language processing. However, as these networks substantially grow in scale, ensuring efficient performance across the entire lifecycle - from cloud-based training to edge-device inference - remains a crucial problem. Our work addresses this need by developing performance modeling and optimization techniques for both cloud-based distributed training and edge-based inference.
First, we develop training throughput prediction models (coarse-grained and fine-grained) for distributed stochastic gradient descent (SGD), characterizing the impact of communication bottlenecks and node stragglers in synchronous/asynchronous and centralized/decentralized settings. Second, we propose an operation-wise framework that accurately predicts the inference latency of various neural architectures - such as CNNs and Vision Transformers (ViTs) - across diverse mobile platforms and ML frameworks. Finally, we propose a heterogeneous co-execution approach that combines low-overhead synchronization with ML-based workload partitioning on mobile CPUs and GPUs, substantially speeding up inference tasks. Together, these three contributions form a comprehensive methodology for end-to-end DNN performance evaluation and optimization, providing practical insights for large-scale training in the cloud and efficient deployment at the edge.Location: Hughes Aircraft Electrical Engineering Center (EEB) - 403
Audiences: Everyone Is Invited
Contact: Zhuojin Li
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
PhD Thesis Proposal - James Hale
Fri, Apr 25, 2025 @ 03:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title of Thesis Proposal:
AI-Mediated Dispute Resolution
Date and Time:
Friday 25 April 2025 3-5PM
Location:
SAL 213
Committee Members:
Dr. Jonathan Gratch, Dr. Gale Lucas, Dr. Jesse Thomason, Dr. Laurent Itti, and Dr. Peter Kim
Abstract:
When conflict arises so does the possibility of potentially irreparable harm interpersonally, policitally, or professionally. Simultaneously, finding effective mediators, especially for those without the means to hire an expert, remains a challenge and may preclude resolution. In this proposal, I examine whether one can leverage recent advances in artificial intelligence to create automated mediators -- democratizing conflict mediation. First, I present a laboratory setting wherein we induce conflict in dyads of human crowd workers as they roleplay a buyer-seller dispute -- yielding the KODIS corpus. Second, we examine whether LLMs can understand emotion dynamics in KODIS to forecast dispute outcomes -- showing they can predict subjective outcomes, and uncovering escalatory spirals as the literature predicts. Lastly, I outline my plan to create automated mediators over the remainder of my PhD.
Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: James Hale
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.
