Konstantin Mishchenko

Latest

• Random Reshuffling: Simple Analysis with Vast Improvements
• Dualize, Split, Randomize: Fast Nonsmooth Optimization Algorithms
• Stochastic Newton and Cubic Newton Methods with Simple Local Linear-Quadratic Rates
• Adaptive Gradient Descent without Descent
• Sinkhorn Algorithm as a Special Case of Stochastic Mirror Descent
• First Analysis of Local GD on Heterogeneous Data
• Tighter Theory for Local SGD on Identical and Heterogeneous Data
• A Self-supervised Approach to Hierarchical Forecasting with Applications to Groupwise Synthetic Controls
• MISO is Making a Comeback With Better Proofs and Rates
• DAve-QN: A Distributed Averaged Quasi-Newton Method with Local Superlinear Convergence Rate
• A Stochastic Decoupling Method for Minimizing the Sum of Smooth and Non-Smooth Functions
• Revisiting Stochastic Extragradient
• Stochastic Distributed Learning with Gradient Quantization and Variance Reduction
• 99% of Worker-Master Communication in Distributed Optimization Is Not Needed
• Distributed Learning with Compressed Gradient Differences
• A Stochastic Penalty Model for Convex and Nonconvex Optimization with Big Constraints
• SEGA: Variance Reduction via Gradient Sketching
• A Delay-tolerant Proximal-Gradient Algorithm for Distributed Learning
• A Distributed Flexible Delay-tolerant Proximal Gradient Algorithm