convex optimization
This branch of optimization focuses on problems where the objective function is convex, meaning any local minimum is also a global minimum. It is prevalent in AI, especially in training algorithms, as it ensures efficient finding of optimal parameters in various models.
- Accelerated Distance-adaptive Methods for Hölder Smooth and Convex Optimization
- AdaLRS: Loss-Guided Adaptive Learning Rate Search for Efficient Foundation Model Pretraining
- Approximate Gradient Coding for Distributed Learning with Heterogeneous Stragglers
- Controlling the Flow: Stability and Convergence for Stochastic Gradient Descent with Decaying Regularization
- Convex Approximation of Two-Layer ReLU Networks for Hidden State Differential Privacy
- Exploring Landscapes for Better Minima along Valleys
- Fast exact recovery of noisy matrix from few entries: the infinity norm approach
- Faster Algorithms for Structured John Ellipsoid Computation
- Gradient-Variation Online Adaptivity for Accelerated Optimization with Hölder Smoothness
- Local Curvature Descent: Squeezing More Curvature out of Standard and Polyak Gradient Descent
- Max Entropy Moment Kalman Filter for Polynomial Systems with Arbitrary Noise
- New Perspectives on the Polyak Stepsize: Surrogate Functions and Negative Results
- Quasi-Self-Concordant Optimization with $\ell_{\infty}$ Lewis Weights
- SpectraLDS: Provable Distillation for Linear Dynamical Systems
- Tight Generalization Bounds for Large-Margin Halfspaces
- Uniform Wrappers: Bridging Concave to Quadratizable Functions in Online Optimization