Variance Reduction

Variance reduction techniques aim to improve the efficiency and stability of various machine learning and optimization algorithms by decreasing the variability in estimates of gradients or other key quantities. Current research focuses on applying these techniques to diverse areas, including meta-learning, reinforcement learning, federated learning, and Monte Carlo methods, often employing algorithms like stochastic variance reduced gradient (SVRG), control variates, and importance sampling. These advancements lead to faster convergence, improved sample efficiency, and more robust performance in various applications, ultimately impacting the scalability and reliability of machine learning models and simulations.

Papers

May 31, 2022

A Computation and Communication Efficient Method for Distributed Nonconvex Problems in the Partial Participation Setting
Alexander Tyurin, Peter Richtárik
Optimization Purpose Stochastic Gradient Variance Reduction Communication Efficient Communication Complexity Client Participation Oracle Complexity Nonconvex Nonconcave Problem

May 29, 2022

Stochastic Zeroth Order Gradient and Hessian Estimators: Variance Reduction and Refined Bias Bounds
Yasong Feng, Tianyu Wang
Hessian Matrix Variance Reduction Zeroth Order Gradient Estimator Optimal Estimation

May 22, 2022

Policy-based Primal-Dual Methods for Concave CMDP with Variance Reduction
Donghao Ying, Mengzi Amy Guo, Hyunin Lee, Yuhao Ding, Javad Lavaei, Zuo-Jun Max Shen
Primal Dual Variance Reduction Global Optimality Concave Function

May 10, 2022

Efficient Burst Raw Denoising with Variance Stabilization and Multi-frequency Denoising Network
Dasong Li, Yi Zhang, Ka Lung Law, Xiaogang Wang, Hongwei Qin, Hongsheng Li
Variance Reduction Denoising Network Multi Frame Denoising Burst Denoising

May 8, 2022

Federated Random Reshuffling with Compression and Variance Reduction
Grigory Malinovsky, Peter Richtárik
Stochastic Gradient Descent Linear Compression Variance Reduction Compression Operator Random Reshuffling

May 6, 2022

Variance Reduction based Partial Trajectory Reuse to Accelerate Policy Gradient Optimization
Hua Zheng, Wei Xie
Policy Gradient Policy Optimization Variance Reduction Trajectory Refinement

April 28, 2022

An Adaptive Incremental Gradient Method With Support for Non-Euclidean Norms
Binghui Xie, Chenhan Jin, Kaiwen Zhou, James Cheng, Wei Meng
Customer Service Variance Reduction Non Euclidean Fast Convergence Large Scale Optimization Incremental Gradient

March 30, 2022

Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization
Yuri Kinoshita, Taiji Suzuki
Application Proficiency Optimization Purpose Langevin Dynamic Convergence Rate Variance Reduction Non Convex Optimization Gradient Complexity

March 22, 2022

Convolutional Neural Network to Restore Low-Dose Digital Breast Tomosynthesis Projections in a Variance Stabilization Domain
Rodrigo de Barros Vimieiro, Chuang Niu, Hongming Shan, Lucas Rodrigues Borges, Ge Wang, Marcelo Andrade da Costa Vieira
Convolutional Neural Network Variance Reduction Polyp Phantom Model Based Deep Learning Digital Breast Tomosynthesis

March 17, 2022

Stochastic Halpern Iteration with Variance Reduction for Stochastic Monotone Inclusions
Xufeng Cai, Chaobing Song, Cristóbal Guzmán, Jelena Diakonikolas
Variance Reduction Monotone Inclusion Stochastic Operator Halpern Iteration

February 21, 2022

MSTGD:A Memory Stochastic sTratified Gradient Descent Method with an Exponential Convergence Rate
Aixiang, Chen, Jinting Zhang, Zanbo Zhang, Zhihong Li
Natural Gradient Variance Reduction Exponential Convergence Rate Sample Gradient Stratified Sampling Gradient Covariance

February 19, 2022

Gradient Estimation with Discrete Stein Operators
Jiaxin Shi, Yuhao Zhou, Jessica Hwang, Michalis K. Titsias, Lester Mackey
Variance Reduction Gradient Estimation Gradient Estimator Control Variate

January 31, 2022

A framework for bilevel optimization that enables stochastic and global variance reduction algorithms
Mathieu Dagréou, Pierre Ablin, Samuel Vaiter, Thomas Moreau
New Framework Bilevel Optimization Empirical Risk Minimization Variance Reduction

January 28, 2022

Adaptive Accelerated (Extra-)Gradient Methods with Variance Reduction
Zijian Liu, Ta Duy Nguyen, Alina Ene, Huy L. Nguyen
Variance Reduction Finite Sum Variance Reduced Adaptive Variance Gradient Evaluation Extra Gradient Method

January 21, 2022

LRSVRG-IMC: An SVRG-Based Algorithm for LowRank Inductive Matrix Completion
Shangrong Yu, Yuxin Chen, Hejun Wu
Sample Complexity Variance Reduction L SVRG Inductive Matrix Completion

December 20, 2021

Decentralized Stochastic Proximal Gradient Descent with Variance Reduction over Time-varying Networks
Xuanjie Li, Yuedong Xu, Jessie Hui Wang, Xin Wang, John C. S. Lui
Stochastic Gradient Decentralized Learning Variance Reduction Proximal Gradient Time Varying Network Decentralized Stochastic Decentralized Proximal

December 14, 2021

Biased Gradient Estimate with Drastic Variance Reduction for Meta Reinforcement Learning
Yunhao Tang
Meta Reinforcement Learning Variance Reduction Gradient Estimation Score Function Unbiased Gradient Biased Gradient

December 5, 2021

Training Structured Neural Networks Through Manifold Identification and Variance Reduction
Zih-Syuan Huang, Ching-pei Lee
Regularization Model Variance Reduction Structured Sparsity Proximal Gradient Descent Non Linear Manifold Unstructured Sparsity Structured Neural Network

November 25, 2021

Distributed Policy Gradient with Variance Reduction in Multi-Agent Reinforcement Learning
Xiaoxiao Zhao, Jinlong Lei, Li Li, Jie Chen
Multi Agent Reinforcement Learning Policy Gradient Optimal Policy Stochastic Optimization Variance Reduction Multi Agent Policy Performance Function

November 23, 2021

Variance Reduction in Deep Learning: More Momentum is All You Need
Lionel Tondji, Sergii Kashubin, Moustapha Cisse
Deep Learning Stochastic Gradient Descent Variance Reduction Variance Aware Large Scale Deep Learning

Variance Reduction

Papers

A Computation and Communication Efficient Method for Distributed Nonconvex Problems in the Partial Participation Setting

Stochastic Zeroth Order Gradient and Hessian Estimators: Variance Reduction and Refined Bias Bounds

Policy-based Primal-Dual Methods for Concave CMDP with Variance Reduction

Efficient Burst Raw Denoising with Variance Stabilization and Multi-frequency Denoising Network

Federated Random Reshuffling with Compression and Variance Reduction

Variance Reduction based Partial Trajectory Reuse to Accelerate Policy Gradient Optimization

An Adaptive Incremental Gradient Method With Support for Non-Euclidean Norms

Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization

Convolutional Neural Network to Restore Low-Dose Digital Breast Tomosynthesis Projections in a Variance Stabilization Domain

Stochastic Halpern Iteration with Variance Reduction for Stochastic Monotone Inclusions

MSTGD:A Memory Stochastic sTratified Gradient Descent Method with an Exponential Convergence Rate

Gradient Estimation with Discrete Stein Operators

A framework for bilevel optimization that enables stochastic and global variance reduction algorithms

Adaptive Accelerated (Extra-)Gradient Methods with Variance Reduction

LRSVRG-IMC: An SVRG-Based Algorithm for LowRank Inductive Matrix Completion

Decentralized Stochastic Proximal Gradient Descent with Variance Reduction over Time-varying Networks

Biased Gradient Estimate with Drastic Variance Reduction for Meta Reinforcement Learning

Training Structured Neural Networks Through Manifold Identification and Variance Reduction

Distributed Policy Gradient with Variance Reduction in Multi-Agent Reinforcement Learning

Variance Reduction in Deep Learning: More Momentum is All You Need