Select a calendar:
Filter March Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for March
-
CS Colloquium: Stephen Tu (USC / ECE) - On the Effectiveness of Generative Modeling for Planning and Control
Wed, Mar 12, 2025 @ 10:00 AM - 11:00 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Stephen Tu, USC / ECE
Talk Title: On the Effectiveness of Generative Modeling for Planning and Control
Abstract: Recent work has demonstrated that modern generative models—including diffusion models and flow matching methods—are a powerful tool for both representing control policies and also designing planning and control algorithms. However, despite strong empirical results, there is a lack of rigorous understanding for why these models work so well in very high-dimensional, autoregressive settings, and surprisingly do not seem to suffer from classic “curse of dimensionality” sample complexity barriers. In this talk, we will shed some light on this phenomenon. First, we will show that shallow diffusion networks can be sample-efficiently learned in the presence of simple latent low-dimensional structures: the intrinsic dimension of the underlying distribution governs the sample complexity, rather than the ambient dimensionality of the problem. Second, we will show that diffusion/flow-matching models and losses are not necessary for learning performant policies in control tasks, and we can actually achieve similar performance using classic energy-based models trained with ranking noise-contrastive estimation—the latter which we prove is nearly asymptotically optimal. We will conclude with some exciting future directions for further investigation into the interplay between generative modeling, controls, and learning.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Stephen Tu is an assistant professor in the Department of Electrical and Computer Engineering at the University of Southern California, where he leads the Statistical Learning for Dynamics and Control group. His research interests span statistical learning theory, safe and optimal control, and robot learning. More specifically, his work has focused on non-asymptotic guarantees for learning dynamical systems, rigorous analysis of distribution shift in feedback settings, safe control synthesis, and more recently foundations of generative modeling. Stephen Tu earned his Ph.D. in Electrical Engineering and Computer Sciences (EECS) from the University of California, Berkeley. Previous to joining USC, Stephen Tu was a research scientist at Google DeepMind Robotics where he focused on combining learning and control-theoretic approaches for robotics.
Host: CS Department
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone (USC) is invited
Contact: CS Faculty Affairs
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.
-
The USC Symposium on the Future of Computing: A 25-Year Vision
Thu, Mar 13, 2025 @ 08:00 AM - 04:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Various, Various
Talk Title: The USC Symposium on the Future of Computing: A 25-Year Vision
Abstract: Join us for the USC Symposium on the Future of Computing: A 25-Year Vision, presented by the Ming Hsieh Department of Electrical & Computer Engineering and the Thomas Lord Department of Computer Science, USC School of Advanced Computing.This two-day event will showcase groundbreaking fundamental and applied research shaping the future of computing over the next quarter century.Featuring leading minds from academia and industry, the symposium will offer keynote and technical sessions spanning a wide range of pivotal topics, including hardware, software, AI and machine learning, theory, and human-computer interaction.Registration is required includes access to the symposium, as well as a light breakfast, lunch, and coffee breaks.RSVP LINK (coming soon)
Biography: DAY 1 & 2 | 8:00am – 4:30pm
- Registration/Check-in
- Keynote Address
- Session A
- Break
- Session B
- Lunch
- Session C
PRESENTATIONS
Day 1:
KEYNOTE SPEAKER
- Amin Vahdat, Google – Engineering Fellow and Vice President for Machine Learning, Systems, and Cloud AI Team
SESSIONS
- AI/ML: Core AI, vision, graphics, robotics
- Hardware II: Processing, architecture, storage for cloud and edge
- Software: OS/networks, databases, programming languages
Day 2:
KEYNOTE SPEAKER
- Doina Precup, McGill University – Professor and Canada Institute for Advanced Research AI Chair
SESSIONS
- Hardware I: Novel computing, quantum technologies, devices
- Human/Computer Interaction: Human in the loop, brain/computer interfaces, edge device interfaces
- Theory: Complexity, algorithms, ML theory, optimization, control, information
Host: Prof. Ramesh Govindan & Prof. Massoud Pedram
Location: Ginsburg Hall (GCS) - Auditorium (LL1)
Audiences: Everyone Is Invited
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. -
CS Colloquium: Yingying (Samara) Ren (ISTA) - Computational Homogenization for Inverse Design of Surface-based Inflatables
Mon, Mar 24, 2025 @ 10:00 AM - 11:00 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Yingying (Samara) Ren, ISTA
Talk Title: Computational Homogenization for Inverse Design of Surface-based Inflatables
Abstract: Surface-based inflatables consist of two nearly inextensible sheets joined along carefully chosen fusing curves, restricting expansion and inducing in-plane contraction and metric frustration. When inflated, these structures settle into a 3D equilibrium that balances elastic and pressure potential energy. In this talk, I will present our computational framework for analyzing and designing such inflatables to approximate a wide range of freeform surfaces while maintaining structural stability. Using numerical homogenization, we characterize periodic inflatable patches with arbitrary fusing patterns and seamlessly combine them through custom nonlinear surface parametrization methods. I will also discuss exciting future research directions in this space.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Yingying Ren (Samara) is an Assitant Professor at ISTA and leads the Geometric Computing and Digital Fabrication group. Her group focuses on research in physics-based simulation, digital fabrication, and computational inverse design. By developing geometric abstractions and efficient numerical methods, her group aims to create new structures and materials with applications in medical devices, architecture, soft robotics, and more.
Host: Oded Stein
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone (USC) is invited
Contact: CS Faculty Affairs
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. -
USC CAIS Seminar
Wed, Mar 26, 2025 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science, USC School of Advanced Computing
Conferences, Lectures, & Seminars
Speaker: Dr. Lily Xu, Oxford and Columbia University
Talk Title: Sequential planning with messy data: RL and restless bandits for planetary health
Abstract: Our planet faces growing crises including biodiversity loss, with animal population sizes declining by 70% since 1970, and maternal mortality, with 1 in 49 girls in low-income countries dying from complications in pregnancy or birth. Underlying these global challenges is the urgent need to effectively allocate scarce resources, often in dynamic environments with limited data. Many of these challenges can be modeled as restless bandits, which traditionally require a perfect model of the environment and relatively small problem sizes. We’ll explore how online learning, deep reinforcement learning, and mixed-integer programming can help overcome these challenges of missing data and complexity.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Lily Xu develops methods across machine learning, optimization, and causal inference for planetary health challenges, with a focus on biodiversity conservation. She aims to enable practitioners to make effective decisions in the face of limited data, taking actions that are robust to uncertainty, effective at scale, and future-looking. In her work, Lily partners closely with NGOs to bridge research and practice, serving as AI Lead for the SMART Partnership. Since 2020, she has co-organized the EAAMO research initiative, committed to advancing Equity and Access in Algorithms, Mechanisms, and Optimization. Lily is currently a postdoctoral fellow at the University of Oxford, with the Leverhulme Centre for Nature Recovery, and will join Columbia IEOR as an Assistant Professor in July 2025. Her research has been recognized with best paper runner-up at AAAI, the INFORMS Doing Good with Good OR award, a Google PhD Fellowship, a Siebel Scholarship, and AAMAS dissertation award runner-up.
Host: Bistra Dilkina
More Info: https://cais.usc.edu/events/sequential-planning-with-messy-data-rl-and-restless-bandits-for-planetary-health/
Location: Ginsburg Hall (GCS) - 107
Audiences: Everyone Is Invited
Contact: Hailey Nadel/USC CAIS
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.
