ICNONLA 2021, Linearly-Constrained Nonsmooth Optimization for Training Deep Neural Network, Chengdu, China, September 2021. SLIDE
SIAM OP23, An Inexact Augmented Lagrangian Algorithm for Training Leaky ReLU Neural Network with Group Sparsity, Seattle, US, June 2023. SLIDE.
Organized Session: Methods for constrained nonconvex problems with application to machine learning, with Ziyu He, and Chuan He.
MOPTA 2023. First-order Methods for Affinely Constrained Composite Non-convex Non-smooth Problems: Lower Complexity Bound and Near-optimal Methods, Lehigh University, US, August 2023. SLIDE
INFORMS Annual meeting 2023. Decentralized Methods for Composite Nonconvex Strongly-Concave Minimax Problems. Phoenix, US, October 2023.
SIAM-NNP 2023. Inexact Proximal Augmented Lagrangian Method for weakly-Convex Problems with Convex Constraints. October 2023. SLIDE.
Organized Session: Large-Scale Structured Optimization, with Mateo Diaz, Benjamin Grimmer, Jiaming Liang, and Yangyang Xu.
INFORMS Conference on Optimization 2024. Lower Complexity Bounds of First-order Methods for Affinely Constrained Composite Nonconvex Problems. Rice University, US, 2024.
Organized Session: Session SunC 4, with Negar, Soheili.
ISMP 2024. A Near-optimal Method for Quadratic Constrained Composite Non-convex Non-smooth Problems, Montréal, Canada, July, 2024.
Organized Session: Nonsmooth nonconvex optimization with geometric structure, with Douglas Soares Goncalves, and Quentin Rebjock.
ICCOPT 2025. A single-loop SPIDER-type stochastic subgradient method for expectation-constrained nonconvex nonsmooth optimization, The University of Southern California (University Park Main Campus), CA, US, July, 2025.
Beihang University, First-order Methods for Affinely Constrained Composite Non-convex Non-smooth Problems: Lower Complexity Bound and Near-optimal Methods, October 2024. SLIDE
PolyU, A stochastic smoothing framework for nonconvex-nonconcave min-sum-max problems with applications to Wasserstein distributionally robust optimization, April 2025.