convergence rates
The speed at which an iterative algorithm approaches its solution, an important metric for evaluating algorithm performance in optimization.
- ASGO: Adaptive Structured Gradient Optimization
- Affine-Invariant Global Non-Asymptotic Convergence Analysis of BFGS under Self-Concordance
- Approximation and Generalization Abilities of Score-based Neural Network Generative Models for Sub-Gaussian Distributions
- Assessing the quality of denoising diffusion models in Wasserstein distance: noisy score and optimal bounds
- Beyond Scores: Proximal Diffusion Models
- Conditional Gradient Methods with Standard LMO for Stochastic Simple Bilevel Optimization
- Continuous-time Riemannian SGD and SVRG Flows on Wasserstein Probabilistic Space
- Convergence Rates for Gradient Descent on the Edge of Stability for Overparametrised Least Squares
- Convergence Rates of Constrained Expected Improvement
- Doubly-Robust Estimation of Counterfactual Policy Mean Embeddings
- Efficient Adaptive Experimentation with Noncompliance
- Efficient Adaptive Federated Optimization
- Efficient Quadratic Corrections for Frank-Wolfe Algorithms
- Establishing Linear Surrogate Regret Bounds for Convex Smooth Losses via Convolutional Fenchel–Young Losses
- Exact and Linear Convergence for Federated Learning under Arbitrary Client Participation is Attainable
- Faster Fixed-Point Methods for Multichain MDPs
- Finite Sample Analysis of Linear Temporal Difference Learning with Arbitrary Features
- From Average-Iterate to Last-Iterate Convergence in Games: A Reduction and Its Applications
- Hamiltonian Descent Algorithms for Optimization: Accelerated Rates via Randomized Integration Time
- Implicit Bias of Spectral Descent and Muon on Multiclass Separable Data
- Least squares variational inference
- Local Curvature Descent: Squeezing More Curvature out of Standard and Polyak Gradient Descent
- Multiclass Loss Geometry Matters for Generalization of Gradient Descent in Separable Classification
- Near-Exponential Savings for Population Mean Estimation with Active Learning
- On Minimax Estimation of Parameters in Softmax-Contaminated Mixture of Experts
- On the Convergence of Single-Timescale Actor-Critic
- Online robust locally differentially private learning for nonparametric regression
- PseuZO: Pseudo-Zeroth-Order Algorithm for Training Deep Neural Networks
- Regularized least squares learning with heavy-tailed noise is minimax optimal
- Risk Bounds For Distributional Regression
- SUMO: Subspace-Aware Moment-Orthogonalization for Accelerating Memory-Efficient LLM Training
- Sample Complexity of Distributionally Robust Average-Reward Reinforcement Learning
- Simple and Optimal Sublinear Algorithms for Mean Estimation
- Small Resamples, Sharp Guarantees: Convergence Rates for Resampled Studentized Quantile Estimators
- Stability and Sharper Risk Bounds with Convergence Rate $\tilde{O}(1/n^2)$
- Stable Minima of ReLU Neural Networks Suffer from the Curse of Dimensionality: The Neural Shattering Phenomenon
- Statistical inference for Linear Stochastic Approximation with Markovian Noise
- Stochastic Gradients under Nuisances
- Zeroth-Order Optimization Finds Flat Minima