Class Weight

Class weighting is a technique used in machine learning to address class imbalance, where some categories in a dataset are significantly under-represented compared to others. Current research focuses on developing effective class weighting strategies within various machine learning models, including those used for image segmentation, object detection, and speech recognition, often incorporating gradient-based methods or integrating class weighting with other techniques like adversarial learning or focal loss. These advancements aim to improve model performance on minority classes, leading to more robust and equitable predictions across all categories, with applications ranging from medical image analysis to natural language processing. The ultimate goal is to develop more reliable and generalizable machine learning models that are less susceptible to bias introduced by imbalanced data.

Papers

September 29, 2024

Balancing the Scales: A Comprehensive Study on Tackling Class Imbalance in Binary Classification
Mohamed Abdelhamid, Abhyuday Desai
Class Imbalance Binary Classification Comprehensive Study Ghanaian Musical Scale Synthetic Minority Oversampling Technique Class Weight

July 1, 2024

Gradient-based Class Weighting for Unsupervised Domain Adaptation in Dense Prediction Visual Tasks
Roberto Alcover-Couso, Marcos Escudero-Viñolo, Juan C. SanMiguel, Jesus Bescós
Domain Adaptation Unsupervised Domain Adaptation Class Imbalance Loss Gradient Class Specific Feature Class Weight Dense Visual Prediction

May 30, 2024

1st Place Solution to Odyssey Emotion Recognition Challenge Task1: Tackling Class Imbalance Problem
Mingjie Chen, Hezhao Zhang, Yuanchao Li, Jiachen Luo, Wen Wu, Ziyang Ma, Peter Bell, Catherine Lai, Joshua Reiss, Lin Wang, Philip C. Woodland, Xie Chen, Huy Phan, Thomas Hain
Speech Emotion Recognition Class Imbalance Place Solution Emotional Speech Class Weight

March 28, 2024

CAT: Exploiting Inter-Class Dynamics for Domain Adaptive Object Detection
Mikhail Kennerley, Jian-Gang Wang, Bharadwaj Veeravalli, Robby T. Tan
Domain Adaptation Domain Adaptive Object Detection Adaptive Object Detection Intra Class Class Weight Semi Supervised Teacher Student Multiple Teacher

January 22, 2024

SFC: Shared Feature Calibration in Weakly Supervised Semantic Segmentation
Xinqiao Zhao, Feilong Tang, Xiaoyang Wang, Jimin Xiao
Semantic Segmentation Semantic Segmentation Model Weakly Supervised Semantic Segmentation Class Activation Class Weight Feature Calibration

December 3, 2022

Leveraging Angular Information Between Feature and Classifier for Long-tailed Learning: A Prediction Reformulation Approach
Haoxuan Wang, Junchi Yan
Simple Classifier Long Tailed Content Based Feature Long Tailed Learning ImageNet LT Exploratory Reformulation Class Weight Angle Prediction

October 25, 2022

FedClassAvg: Local Representation Learning for Personalized Federated Learning on Heterogeneous Neural Networks
Jaehee Jang, Heonseok Ha, Dahuin Jung, Sungroh Yoon
Deep Neural Network Personalized Federated Learning FedAvg Converges Class Weight Classification Layer Ensemble Averaging Heterogeneous Autoencoder

July 17, 2022

Coupling Adversarial Learning with Selective Voting Strategy for Distribution Alignment in Partial Domain Adaptation
Sandipan Choudhuri, Hemanth Venkateswara, Arunabha Sen
Domain Invariant Feature Distribution Alignment Voting Method Open Set Domain Adaptation Class Weight Partial Domain Adaptation Cross Domain Classification Weight Coupling Learning

May 19, 2022

Why only Micro-F1? Class Weighting of Measures for Relation Classification
David Harbecke, Yuxuan Chen, Leonhard Hennig, Christoph Alt
Imbalanced Datasets New Measure Calysto Scheme Relation Classification Class Weight Weighting Scheme

May 10, 2022

Calibrating for Class Weights by Modeling Machine Learning
Andrew Caplin, Daniel Martin, Philip Marx
Machine Learning Class Imbalance Calibration Performance Machine Learning Algorithm Pneumonia Detection Class Weight

April 13, 2022

Calibrating Class Weights with Multi-Modal Information for Partial Video Domain Adaptation
Xiyu Wang, Yuecong Xu, Kezhi Mao, Jianfei Yang
Cross Domain Multimodal Information Video Domain Adaptation Class Calibration Class Weight Calibration Attack

April 12, 2022

On the Equity of Nuclear Norm Maximization in Unsupervised Domain Adaptation
Wenju Zhang, Xiang Zhang, Qing Liao, Long Lan, Mengzhu Wang, Wei Wang, Baoyun Peng, Zhengming Ding
Domain Adaptation Unsupervised Domain Adaptation Cross Domain Nuclear Norm Class Weight

February 9, 2022

Class Distance Weighted Cross-Entropy Loss for Ulcerative Colitis Severity Estimation
Gorkem Polat, Ilkay Ergenc, Haluk Tarik Kani, Yesim Ozen Alahdab, Ozlen Atug, Alptekin Temizel
Loss Function Cross Entropy Loss Class Weight Categorical Cross Entropy

December 14, 2021

E-CRF: Embedded Conditional Random Field for Boundary-caused Class Weights Confusion in Semantic Segmentation
Jie Zhu, Huabin Huang, Banghuai Li, Leye Wang
Semantic Segmentation Segmentation Model Segmentation Performance Conditional Random Field Fully Connected CRFs Class Weight

December 1, 2021

Learning from Mistakes based on Class Weighting with Application to Neural Architecture Search
Jay Gala, Pengtao Xie
LeArning Abstract Application Proficiency Neural Architecture Search Human Answer Mistake Learning With Error Class Weight Multi Level Optimization Class Conditional Impression Reappearing

Class Weight

Papers

Balancing the Scales: A Comprehensive Study on Tackling Class Imbalance in Binary Classification

Gradient-based Class Weighting for Unsupervised Domain Adaptation in Dense Prediction Visual Tasks

1st Place Solution to Odyssey Emotion Recognition Challenge Task1: Tackling Class Imbalance Problem

CAT: Exploiting Inter-Class Dynamics for Domain Adaptive Object Detection

SFC: Shared Feature Calibration in Weakly Supervised Semantic Segmentation

Leveraging Angular Information Between Feature and Classifier for Long-tailed Learning: A Prediction Reformulation Approach

FedClassAvg: Local Representation Learning for Personalized Federated Learning on Heterogeneous Neural Networks

Coupling Adversarial Learning with Selective Voting Strategy for Distribution Alignment in Partial Domain Adaptation

Why only Micro-F1? Class Weighting of Measures for Relation Classification

Calibrating for Class Weights by Modeling Machine Learning

Calibrating Class Weights with Multi-Modal Information for Partial Video Domain Adaptation

On the Equity of Nuclear Norm Maximization in Unsupervised Domain Adaptation

Class Distance Weighted Cross-Entropy Loss for Ulcerative Colitis Severity Estimation

E-CRF: Embedded Conditional Random Field for Boundary-caused Class Weights Confusion in Semantic Segmentation

Learning from Mistakes based on Class Weighting with Application to Neural Architecture Search