Content tagged with ICML

• [C105] What preferences can–and cannot–predict in multi-agent online learning
• [C104] Taming stochastic gradient descent: Almost sure convergence and saddle-point avoidance under (L0,L1)-smoothness
• [C103] Bregman meets Lévy: Stochastic mirror descent with heavy-tailed noise in continuous and discrete time
• [C97] The impact of uncertainty on regularized learning in games
• [C96] The global convergence time of stochastic gradient descent in non-convex landscapes: Sharp estimates via large deviations
• [C92] A geometric decomposition of finite games: Convergence vs. recurrence under exponential weights
• [C91] The computational complexity of finding second-order stationary points
• [C90] What is the long-run distribution of stochastic gradient descent? A large deviations analysis
• [C78] AdaGrad avoids saddle points
• [C77] UnderGrad: A universal black-box optimization method with almost dimension-free convergence rate guarantees
• [C76] Nested bandits
• [C64] The limits of min-max optimization algorithms: Convergence to spurious non-critical sets
• [C63] Zeroth-order non-convex learning via hierarchical dual averaging
• [C62] Regret minimization in stochastic non-convex learning via a proximal-gradient approach
• [C56] A new regret analysis for Adam-type algorithms
• [C55] Gradient-free online learning in continuous games with delayed rewards
• [C54] Finite-time last-iterate convergence for multi-agent learning in games
• [C46] Cautious regret minimization: Online optimization with long-term budget constraints
• [C37] Distributed asynchronous optimization with unbounded delays: How slow can you go?