Aryan Mokhtari

23

Papers

65

Total Citations

Papers (23)

Adaptive Newton Method for Empirical Risk Minimization to Statistical Accuracy

NeurIPS 2016arXiv

First-Order Adaptive Sample Size Methods to Reduce Complexity of Empirical Risk Minimization

NeurIPS 2017arXiv

On the Crucial Role of Initialization for Matrix Factorization

Learning Mixtures of Experts with EM: A Mirror Descent Perspective

Affine-Invariant Global Non-Asymptotic Convergence Analysis of BFGS under Self-Concordance

NeurIPS 2025arXiv

Provable Multi-Task Representation Learning by Two-Layer ReLU Neural Networks

Submodular Meta-Learning

Task-Robust Model-Agnostic Meta-Learning

Second Order Optimality in Decentralized Non-Convex Optimization via Perturbed Gradient Tracking

On the Convergence Theory of Debiased Model-Agnostic Meta-Reinforcement Learning

Exploiting Local Convergence of Quasi-Newton Methods Globally: Adaptive Sample Size Approach

Generalization of Model-Agnostic Meta-Learning Algorithms: Recurring and Unseen Tasks

FedAvg with Fine Tuning: Local Updates Lead to Representation Learning

Projection-Free Methods for Stochastic Simple Bilevel Optimization with Convex Lower-level Problem

Accelerated Quasi-Newton Proximal Extragradient: Faster Rate for Smooth Convex Optimization

Greedy Pruning with Group Lasso Provably Generalizes for Matrix Sensing

Decentralized Submodular Maximization: Bridging Discrete and Continuous Settings

Towards More Efficient Stochastic Decentralized Learning: Faster Convergence and Sparse Communication

Escaping Saddle Points in Constrained Optimization

Direct Runge-Kutta Discretization Achieves Acceleration

Robust and Communication-Efficient Collaborative Learning

Stochastic Continuous Greedy ++: When Upper and Lower Bounds Match

Personalized Federated Learning with Theoretical Guarantees: A Model-Agnostic Meta-Learning Approach