Class Wise Multiplier

Class-wise multipliers are parameters used to adjust the contribution of individual classes within machine learning models, particularly in addressing class imbalance or improving calibration. Research focuses on incorporating these multipliers into optimization algorithms like the Alternating Direction Method of Multipliers (ADMM) and employing them within various architectures, including deep neural networks and recurrent neural networks, to enhance model performance and efficiency. This approach offers significant potential for improving the accuracy and robustness of machine learning models across diverse applications, particularly in scenarios with skewed class distributions or where specific classes require more careful weighting.

Papers

November 9, 2024

Online Parallel Multi-Task Relationship Learning via Alternating Direction Method of Multipliers
Ruiyu Li, Peilin Zhao, Guangxia Li, Zhiqiang Xu, Xuewei Li
Loss Function Gradient Descent Multitask Learning Online Service Class Wise Multiplier Parallel Optimization

October 2, 2024

Auto-conditioned primal-dual hybrid gradient method and alternating direction method of multipliers
Guanghui Lan, Tianjiao Li
Primal Dual Gradient Method Saddle Point Class Wise Multiplier

August 28, 2024

Alternating Direction Method of Multipliers for Negative Binomial Model with The Weighted Difference of Anisotropic and Isotropic Total Variation
Yu Lu, Kevin Bui, Roummel F. Marcia
Low Light Total Variation Poisson Noise Class Wise Multiplier Weighted Average Regularization Penalty Photon Efficient Imaging

August 9, 2024

Optimal Distributed Multi-Robot Communication-Aware Trajectory Planning using Alternating Direction Method of Multipliers
Jeppe Heini Mikkelsen, Roberto Galeazzi, Matteo Fumagalli
Collision Avoidance Multi Robot System Optimisation Problem Class Wise Multiplier

March 31, 2024

RL-MUL: Multiplier Design Optimization with Deep Reinforcement Learning
Dongsheng Zuo, Jiadong Zhu, Yikang Ouyang, Yuzhe Ma
Reinforcement Learning Deep Reinforcement Learning Class Wise Multiplier Multiply Accumulate Multiplier Design Rl Msa

February 25, 2024

EncodingNet: A Novel Encoding-based MAC Design for Efficient Neural Network Acceleration
Bo Liu, Grace Li Zhang, Xunzhao Yin, Ulf Schlichtmann, Bing Li
Deep Neural Network Long Form Novel Class Wise Multiplier Deep Learning Acceleration Multiply Accumulate Non Uniform Quantization

January 29, 2024

Distributed Markov Chain Monte Carlo Sampling based on the Alternating Direction Method of Multipliers
Alexandros E. Tzikas, Licio Romao, Mert Pilanci, Alessandro Abate, Mykel J. Kochenderfer
Optimization Purpose Convex Optimization Gradient Method Fast Convergence Class Wise Multiplier Local Sampling

January 2, 2024

Evolutionary Alternating Direction Method of Multipliers for Constrained Multi-Objective Optimization with Unknown Constraints
Shuang Li, Ke Li, Wei Li, Ming Yang
Multi Objective Optimization Participation Constraint Evolutionary Computation Evolutionary Multi Objective Optimization Class Wise Multiplier Unknown Constraint Constrained Multi Objective Evolutionary Algorithm Constrained Multi Objective Optimization Problem

August 7, 2023

Almost-sure convergence of iterates and multipliers in stochastic sequential quadratic optimization
Frank E. Curtis, Xin Jiang, Qi Wang
Stochastic Gradient Class Wise Multiplier Sequential Quadratic Programming Almost Sure Convergence

April 23, 2023

Using Alternation Direction Method of Multipliers to Enhance robots Calibration Accuracy based on Multi-Planal Constraints
Tinghui Chen, Shuai Li
Industrial Robot Geometric Constraint Class Wise Multiplier Positioning Accuracy Robot Calibration

April 6, 2023

Low-Latency Online Multiplier with Reduced Activities and Minimized Interconnect for Inner Product Arrays
Muhammad Usman, Milos Ercegovac, Jeong-A Lee
Class Wise Multiplier Inner Product Dataflow Accelerator

March 25, 2023

Multi-agent Black-box Optimization using a Bayesian Approach to Alternating Direction Method of Multipliers
Dinesh Krishnamoorthy, Joel A. Paulson
Bayesian Optimization Multi Agent System Bayesian Approach Multi Agent Network Class Wise Multiplier Black Box Optimization Problem Numerical Black Box Optimization Multi Agent Bayesian Optimization

February 7, 2023

Averaged Method of Multipliers for Bi-Level Optimization without Lower-Level Strong Convexity
Risheng Liu, Yaohua Liu, Wei Yao, Shangzhi Zeng, Jin Zhang
Gradient Norm Gradient Method Bi Level Optimization Strong Convexity Class Wise Multiplier Single Loop Averaging Algorithm

November 28, 2022

Class Adaptive Network Calibration
Bingyuan Liu, Jérôme Rony, Adrian Galdran, Jose Dolz, Ismail Ben Ayed
Calibration Performance Label Smoothing Class Wise Multiplier Class Calibration

October 12, 2022

Optimizing Evaluation Metrics for Multi-Task Learning via the Alternating Direction Method of Multipliers
Ge-Yang Ke, Yan Pan, Jian Yin, Chang-Qin Huang
Multi Task Learning Generalization Performance Evaluation Metric Class Wise Multiplier Hinge Loss

October 8, 2022

Low Error-Rate Approximate Multiplier Design for DNNs with Hardware-Driven Co-Optimization
Yao Lu, Jide Zhang, Su Zheng, Zhen Li, Lingli Wang
DNN Framework Co Design Low Priority Neural Network Weight DNN Accuracy Class Wise Multiplier Approximate Multiplier

May 20, 2022

Towards Extremely Fast Bilevel Optimization with Self-governed Convergence Guarantees
Risheng Liu, Xuan Liu, Wei Yao, Shangzhi Zeng, Jin Zhang
Bilevel Optimization Gradient Method Bi Level Optimization Gradient Approximation Class Wise Multiplier Non Asymptotic Convergence

April 11, 2022

Multiplier with Reduced Activities and Minimized Interconnect for Inner Product Arrays
Muhammad Usman, Jeong-A Lee, Milos D. Ercegovac
Class Wise Multiplier Inner Product Asynchronous Parallel

January 20, 2022

HEAM: High-Efficiency Approximate Multiplier Optimization for Deep Neural Networks
Su Zheng, Zhen Li, Yao Lu, Jingbo Gao, Jide Zhang, Lingli Wang
Deep Neural Network DNN Framework DNN Accelerator Class Wise Multiplier Accuracy Loss Approximate Multiplier

December 31, 2021

Training Recurrent Neural Networks by Sequential Least Squares and the Alternating Direction Method of Multipliers
Alberto Bemporad
Nonlinear Dynamic Sequential Learning Alternating Direction Method of Multiplier Nonlinear System Identification Class Wise Multiplier

Class Wise Multiplier

Papers

Online Parallel Multi-Task Relationship Learning via Alternating Direction Method of Multipliers

Auto-conditioned primal-dual hybrid gradient method and alternating direction method of multipliers

Alternating Direction Method of Multipliers for Negative Binomial Model with The Weighted Difference of Anisotropic and Isotropic Total Variation

Optimal Distributed Multi-Robot Communication-Aware Trajectory Planning using Alternating Direction Method of Multipliers

RL-MUL: Multiplier Design Optimization with Deep Reinforcement Learning

EncodingNet: A Novel Encoding-based MAC Design for Efficient Neural Network Acceleration

Distributed Markov Chain Monte Carlo Sampling based on the Alternating Direction Method of Multipliers

Evolutionary Alternating Direction Method of Multipliers for Constrained Multi-Objective Optimization with Unknown Constraints

Almost-sure convergence of iterates and multipliers in stochastic sequential quadratic optimization

Using Alternation Direction Method of Multipliers to Enhance robots Calibration Accuracy based on Multi-Planal Constraints

Low-Latency Online Multiplier with Reduced Activities and Minimized Interconnect for Inner Product Arrays

Multi-agent Black-box Optimization using a Bayesian Approach to Alternating Direction Method of Multipliers

Averaged Method of Multipliers for Bi-Level Optimization without Lower-Level Strong Convexity

Class Adaptive Network Calibration

Optimizing Evaluation Metrics for Multi-Task Learning via the Alternating Direction Method of Multipliers

Low Error-Rate Approximate Multiplier Design for DNNs with Hardware-Driven Co-Optimization

Towards Extremely Fast Bilevel Optimization with Self-governed Convergence Guarantees

Multiplier with Reduced Activities and Minimized Interconnect for Inner Product Arrays

HEAM: High-Efficiency Approximate Multiplier Optimization for Deep Neural Networks

Training Recurrent Neural Networks by Sequential Least Squares and the Alternating Direction Method of Multipliers