November 19, 2020 @ 4:00 pm - 5:00 pm Learning via early stopping and untrained neural nets Remote/Virtual
December 3, 2020 @ 4:00 pm - 5:00 pm Approximate planning and learning in partially observed systems Remote/Virtual
December 10, 2020 @ 4:00 pm - 5:00 pm Connections between Online Learning and Differential Privacy Remote/Virtual
January 29, 2021 @ 9:30 am - 10:30 am Fastest Identification in Linear Systems, Joint work with Yassir Jedra Remote/Virtual
February 4, 2021 @ 4:00 pm - 5:00 pm Reinforcement Learning for Mean Field Games with Strategic Complementarities Remote/Virtual
February 18, 2021 @ 4:00 pm - 5:00 pm A Lyapunov Theory for Finite-Sample Guarantees of Asynchronous, Value-Based Reinforcement Learning Algorithms Remote/Virtual
March 18, 2021 @ 4:00 pm - 5:00 pm MLE and the EM algorithm for Mixtures: Minimax Results Remote/Virtual
March 25, 2021 @ 4:00 pm - 5:00 pm Communication-Efficient Optimization Methods for Federated Learning Remote/Virtual
April 1, 2021 @ 4:00 pm - 5:00 pm Simple Agent, Complex Environment: Efficient Reinforcement Learning with Agent State Remote/Virtual
April 15, 2021 @ 4:00 pm - 5:00 pm Risk-Sensitive Market Clearing for the Modern Power System Remote/Virtual
April 22, 2021 @ 4:00 pm - 5:00 pm Postponed: Seminar by Aylin Yener, ECE/CSE/IE @ OSU Remote/Virtual
April 22, 2021 @ 4:00 pm - 5:00 pm Adaptive Design of Switchback Experiments: A Markov Chain Perspective Remote/Virtual
September 2, 2021 @ 4:00 pm - 5:00 pm Multiclass and Multi-Label Learning with General Losses: What is the Right Output Coding and Decoding? Remote/Virtual
September 9, 2021 @ 4:00 pm - 5:00 pm How to Schedule Near-Optimally under Real-World Constraints Remote/Virtual
September 16, 2021 @ 4:00 pm - 5:00 pm The Power of Exploiter: Provable Multi-Agent RL in Large State Spaces Remote/Virtual
October 1, 2021 @ 9:30 am - 10:30 am Database Alignment: Fundamental Limits and Efficient Algorithms Remote/Virtual
October 15, 2021 @ 9:30 am - 10:30 am Social Learning in Multi-Agent, Multi-Armed Bandits Remote/Virtual
October 21, 2021 @ 11:00 am - 12:00 pm Communication and Sensing: From Compressed Sampling to Model-based Deep Learning Remote/Virtual
November 4, 2021 @ 4:00 pm - 5:00 pm Spectrum Sharing for 5G and Beyond: A Network Economics View Remote/Virtual
November 18, 2021 @ 12:00 pm - 1:00 pm Finding Global Minima via Kernel Approximations Remote/Virtual
December 2, 2021 @ 4:00 pm - 5:00 pm Generalization and Inductive Bias in Neural Networks Remote/Virtual
December 9, 2021 @ 4:00 pm - 5:00 pm Data-Centric Networking: Theory, Algorithms and Applications Remote/Virtual
January 6, 2022 @ 4:00 pm - 5:00 pm Data-driven and Learning-based Methods for Control: Challenges and New Horizons Remote/Virtual
January 13, 2022 @ 4:00 pm - 5:00 pm Sequential Decision Making: How Much Adaptivity Is Needed Anyways? Remote/Virtual
January 20, 2022 @ 4:00 pm - 5:00 pm Surprises in High-Dimensional (Overparameterized) Linear Classification Remote/Virtual
September 8, 2022 @ 4:00 pm - 5:00 pm Blessing of Nonconvexity in Factorized Models 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
September 15, 2022 @ 4:00 pm - 5:00 pm Principles of Deep Representation Learning via Neural Collapse 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
September 22, 2022 @ 4:00 pm - 5:00 pm Robust Learning by Double Over-Parameterization 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
October 6, 2022 @ 4:00 pm - 5:00 pm A New Embedding Framework to Design Consistent Surrogate Losses 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
October 13, 2022 @ 4:00 pm - 5:00 pm Method of Moments: From Sample Complexity to Efficient Implicit Computations Remote/Virtual
October 27, 2022 @ 4:00 pm - 5:00 pm Accelerating Ill-conditioned Low-rank Estimation via Scaled Gradient Descent 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
November 10, 2022 @ 4:00 pm - 5:00 pm Opportunistic Integrated Communication and Weather Sensing 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
November 17, 2022 @ 4:00 pm - 5:00 pm Advancing Machine Learning for Imaging: Regularization and Robustness 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
December 8, 2022 @ 4:00 pm - 5:00 pm Deep Networks and the Multiple Manifold Problem 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
January 12, 2023 @ 9:00 am - 10:00 pm Dynamic Pricing and Matching for Online Marketplaces 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
February 2, 2023 @ 9:15 am - 10:15 am Reinforcement Learning with Robustness and Safety Guarantees 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
February 9, 2023 @ 9:15 am - 10:15 am Understanding the Universality Phenomenon in High- Dimensional Estimation and Learning: Some Recent Progress 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
February 16, 2023 @ 9:15 am - 10:15 am Evaluating Unsupervised Denoising Requires Unsupervised Metrics Remote/Virtual
February 23, 2023 @ 9:15 am - 10:15 am Cancelled: Implicit Geometry of Large Classification Models: Imbalance Trouble 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
March 16, 2023 @ 9:15 am - 10:15 am Implicit Geometry of Large Classification Models: Imbalance Trouble Remote/Virtual
March 21, 2023 @ 10:00 am - 11:00 am Load Balancing under Data Locality: Extending Mean-Field Framework to Constrained Large-Scale Systems 1012 EECS Building 1301 Beal Ave, Ann Arbor, MI, United States
March 23, 2023 @ 9:15 am - 10:15 am Explainable AI via Semantic Information Pursuit 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
March 30, 2023 @ 9:15 am - 10:15 am Deep Learning Meets Sparse Regularization: A Signal Processing Perspective 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
April 6, 2023 @ 9:15 am - 10:15 am Best of three worlds? Bias and extrapolation in constant-step size stochastic approximation 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
April 20, 2023 @ 9:15 am - 10:15 am Machine Learning challenges in Metrology in Semiconductor Device Industry 1690 Beyster Building 2260 Hayward Street, Ann Arbor, MI, United States
May 8, 2023 @ 3:00 pm - 4:00 pm Demystifying and Mitigating Unfairness for Machine Learning over Graphs 3316 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
August 2, 2023 @ 10:00 am - 11:00 am Setup Times in Multiserver Systems 2311 EECS 1301 Beal Avenue, Ann Arbor, MI, United States
September 7, 2023 @ 3:30 pm - 5:00 pm On the Emergence of Invariant Low-Dimensional Subspaces in Gradient Descent for Learning Deep Networks 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
September 15, 2023 @ 11:00 am - 12:30 pm On an Asymptotic Criterion for Blockchain Design 2311 EECS 1301 Beal Avenue, Ann Arbor, MI, United States
September 21, 2023 @ 3:30 pm - 5:00 pm Quantum Signal Processing 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
September 22, 2023 @ 3:30 pm - 5:00 pm Integrated Sensing and Communication: aspects from Information Theory, Communication Theory, and Machine Learning 1012 EECS Building 1301 Beal Ave, Ann Arbor, MI, United States
September 28, 2023 @ 3:30 pm - 5:00 pm Solving Inverse Problems with Latent Diffusion Models via Hard Data Consistency 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 2, 2023 @ 9:00 am - 10:30 am Wireless is back: The 6G Revolution Towards Connected Intelligence 1500 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 5, 2023 @ 3:30 pm - 5:00 pm Transformers as Support Vector Machines 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 12, 2023 @ 3:30 pm - 5:00 pm Theoretical Characterization of Forgetting and Generalization of Continual Learning 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 19, 2023 @ 3:30 pm - 5:00 pm How Hessian Structure Explains Mysteries in Sharpness Regularization 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 26, 2023 @ 3:30 pm - 5:00 pm Breaking the Sample Size Barrier in Reinforcement Learning via Model-Based Algorithms 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
November 2, 2023 @ 3:30 pm - 5:00 pm Overparameterization and Global Optimality in Nonconvex Low-Rank Matrix Estimation and Optimization 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
November 9, 2023 @ 3:30 pm - 5:00 pm Sparsity for Efficient Long Sequence Generation of LLMs 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
November 16, 2023 @ 3:30 pm - 5:00 pm Rethinking the theoretical foundation of reinforcement learning 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
November 30, 2023 @ 3:30 pm - 5:00 pm (CANCELED) Discrete Optimization for Adversarial Attacks on Large Language Models 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
December 7, 2023 @ 3:30 pm - 5:00 pm Flat Minima Generalize for Low-Rank Matrix Recovery 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
December 8, 2023 @ 3:30 pm - 4:30 pm Towards a Theoretical Understanding of Parameter-Efficient Fine-Tuning (and Beyond) 1303 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
January 18 @ 3:30 pm - 4:30 pm Accelerated Optimization for Dynamic MRI Reconstruction with Locally Low-Rank Regularizers 3433 EECS Building 1301 Beal Ave, Ann Arbor, MI, United States
January 25 @ 3:30 pm - 4:30 pm Transfer learning for contextual multi-armed bandits 3433 EECS Building 1301 Beal Ave, Ann Arbor, MI, United States
January 26 @ 1:30 pm - 2:30 pm Foundation Models AAA: Acceleration, Application, Adversary 3433 EECS Building 1301 Beal Ave, Ann Arbor, MI, United States
February 1 @ 3:30 pm - 4:30 pm On False Positive Error 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
February 8 @ 3:30 pm - 4:30 pm Re-engineering the Uplink for Next Generation Multiple Access 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
February 15 @ 3:30 pm - 4:30 pm Select Topics Associated with 3D Scanning by Means of Structured Light Illumination 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
March 7 @ 3:30 pm - 4:30 pm Formal privacy guarantees for optimization datasets in power systems 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
March 14 @ 3:30 pm - 4:30 pm Image Restoration through Inversion by Direct Iteration (InDI) 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
March 21 @ 3:30 pm - 4:30 pm Distributed learning on correlated data with communication constraints 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
March 28 @ 3:30 pm - 4:30 pm Cancelled 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
April 11 @ 3:30 pm - 4:30 pm Edge and IoT-supported Intelligent Augmented Reality: Promise, Challenges, and Solutions 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
April 18 @ 3:30 pm - 4:30 pm Data-Analytic Opportunities and Challenges in Solar Eruption Forecasting 1200 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
June 7 @ 2:00 pm - 3:00 pm Stochastic Optimisation Methods Applied to Image Reconstruction: results, software and outlook. 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
June 13 @ 4:00 pm - 5:00 pm Novel Technologies for Accelerated MRI 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
August 29 @ 3:30 pm - 4:30 pm Interdependent privacy in third-party apps: past, present, and future 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
September 5 @ 3:30 pm - 4:30 pm Permutation-Free Kernel Hypothesis Tests 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
September 12 @ 3:30 pm - 4:30 pm Computational Imaging through Atmospheric Turbulence 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
September 19 @ 3:30 pm - 4:30 pm A Statistical Framework for Private Personalized Federated Learning and Estimation 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
September 26 @ 3:30 pm - 4:30 pm Understanding Distribution Learning of Diffusion Models via Low-dimensional Modeling 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 3 @ 3:30 pm - 4:30 pm Approximate independence of permutation mixtures 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 10 @ 3:30 pm - 4:30 pm Toward (Truly) Resilient Computing and Communications in 6G Mobile Systems 3427 EECS Building 1301 Beal Avenue, Ann Arbor, MI, United States
October 11 @ 3:30 pm - 4:30 pm Domain Counterfactuals for Explainability, Fairness, and Domain Generalization 1008 EECS Building 1301 Beal Ave, Ann Arbor, United States
MENU
Search
MENU
Search
