Class Distribution

Class distribution, the frequency of different categories within a dataset, is a critical factor influencing the performance and fairness of machine learning models. Current research focuses on mitigating the negative impacts of imbalanced class distributions, particularly in semi-supervised learning and federated learning settings, employing techniques like dual-branch training, class-wise averaging, and adaptive distribution fusion. These advancements aim to improve model accuracy and robustness across all classes, addressing biases towards majority categories and enhancing generalization to unseen data, with significant implications for various applications including medical image analysis, facial expression recognition, and object detection.

Papers

November 2, 2024

Covariance-based Space Regularization for Few-shot Class Incremental Learning
Yijie Hu, Guanyu Yang, Zhaorui Tan, Xiaowei Huang, Kaizhu Huang, Qiu-Feng Wang
LeArning Abstract Discriminative Representation Class Distribution Shot Training Covariance Neural Network Spatially Varying Regularization

October 9, 2024

Calibrating Verbalized Probabilities for Large Language Models
Cheng Wang, Gyuri Szarvas, Georges Balazs, Pavel Danchenko, Patrick Ernst
Softmax Function Relative Probability Discriminative Task Class Distribution Black Box Large Language Model

June 19, 2024

Boosting Consistency in Dual Training for Long-Tailed Semi-Supervised Learning
Kai Gan, Tong Wei, Min-Ling Zhang
Strong Consistency Class Distribution Bias Label Simultaneous Training Long Tailed Semi Supervised Learning

June 12, 2024

Regularizing and Aggregating Clients with Class Distribution for Personalized Federated Learning
Gyuejeong Lee, Daeyoung Choi
Personalized Federated Learning Federated Averaging Neural Network Weight Class Distribution Client Clustering

April 24, 2024

Ada-DF: An Adaptive Label Distribution Fusion Network For Facial Expression Recognition
Shu Liu, Yan Xu, Tongming Wan, Xiaoyan Kui
Facial Expression Recognition Label Distribution Label Distribution Learning Class Distribution Adaptive Fusion

April 14, 2024

FedDistill: Global Model Distillation for Local Model De-Biasing in Non-IID Federated Learning
Changlin Song, Divya Saxena, Jiannong Cao, Yuqing Zhao
Anti Forgetting Mutual Distillation Class Distribution Global Distillation

February 21, 2024

SimPro: A Simple Probabilistic Framework Towards Realistic Long-Tailed Semi-Supervised Learning
Chaoqun Du, Yizeng Han, Gao Huang
Semi Supervised Learning Probabilistic Model Class Distribution Class Distribution Estimation Long Tailed Semi Supervised Learning

February 9, 2024

A Kalman Filter Based Framework for Monitoring the Performance of In-Hospital Mortality Prediction Models Over Time
Jiacheng Liu, Lisa Kirkland, Jaideep Srivastava
System Performance Time Matter Kalman Filter Class Distribution Sample Size Precision Recall Curve Hospital Mortality Prediction

January 29, 2024

Rectify the Regression Bias in Long-Tailed Object Detection
Ke Zhu, Minghao Fu, Jie Shao, Tianyu Liu, Jianxin Wu
Long Tailed Class Distribution Classification Head Prediction Bias Long Tailed Object Detection

December 27, 2023

Twice Class Bias Correction for Imbalanced Semi-Supervised Learning
Lan Li, Bowen Tao, Lu Han, De-chuan Zhan, Han-jia Ye
Semi Supervised Learning Class Imbalance Bias Correction Class Distribution Imbalanced Semi Supervised Learning Class Prior

December 25, 2023

Three Heads Are Better Than One: Complementary Experts for Long-Tailed Semi-supervised Learning
Chengcheng Ma, Ismail Elezi, Jiankang Deng, Weiming Dong, Changsheng Xu
Pseudo Label Batch Normalization Multi Head Class Distribution Complementary Information Long Tailed Semi Supervised Learning

December 21, 2023

Entropic Open-set Active Learning
Bardia Safaei, Vibashan VS, Celso M. de Melo, Vishal M. Patel
Active Learning Stable Entropy Class Distribution Entropy Score

December 9, 2023

Enhancing Robustness of Foundation Model Representations under Provenance-related Distribution Shifts
Xiruo Ding, Zhecheng Sheng, Brian Hur, Feng Chen, Serguei V. S. Pakhomov, Trevor Cohen
Native Robustness Foundation Model Distribution Shift Data Provenance Class Distribution Source Distribution Estimation

December 4, 2023

Generalized Categories Discovery for Long-tailed Recognition
Ziyun Li, Christoph Meinel, Haojin Yang
Long Tailed Recognition Generalized Category Discovery Class Distribution New Category

November 30, 2023

Class Distribution Shifts in Zero-Shot Learning: Learning Robust Representations
Yuli Slavutsky, Yuval Benjamini
Robust Representation Robust Representation Learning Class Distribution Class Distribution Shift

September 25, 2023

FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learning
Dipam Goswami, Yuyang Liu, Bartłomiej Twardowski, Joost van de Weijer
Feature Extractor Prototypical Network Class Distribution Exemplar Free Class Incremental Learning Covariance Alignment Free Continual Learning Frozen Feature

August 17, 2023

Kernel-Based Tests for Likelihood-Free Hypothesis Testing
Patrik Róbert Gerber, Tianze Jiang, Yury Polyanskiy, Rui Sun
Sample Complexity Binary Classification Level Test Class Distribution Non Parametric Likelihood Free Inference

August 16, 2023

August 6, 2023

Novel Class Discovery for Long-tailed Recognition
Chuyu Zhang, Ruijie Xu, Xuming He
Long Tailed Recognition Novel Class Discovery Class Distribution High Quality Benchmark Animal Datasets Self Labeling

Class Distribution

Papers

Covariance-based Space Regularization for Few-shot Class Incremental Learning

Calibrating Verbalized Probabilities for Large Language Models

Boosting Consistency in Dual Training for Long-Tailed Semi-Supervised Learning

Regularizing and Aggregating Clients with Class Distribution for Personalized Federated Learning

Ada-DF: An Adaptive Label Distribution Fusion Network For Facial Expression Recognition

FedDistill: Global Model Distillation for Local Model De-Biasing in Non-IID Federated Learning

SimPro: A Simple Probabilistic Framework Towards Realistic Long-Tailed Semi-Supervised Learning

A Kalman Filter Based Framework for Monitoring the Performance of In-Hospital Mortality Prediction Models Over Time

Rectify the Regression Bias in Long-Tailed Object Detection

Twice Class Bias Correction for Imbalanced Semi-Supervised Learning

Three Heads Are Better Than One: Complementary Experts for Long-Tailed Semi-supervised Learning

Entropic Open-set Active Learning

Enhancing Robustness of Foundation Model Representations under Provenance-related Distribution Shifts

Generalized Categories Discovery for Long-tailed Recognition

Class Distribution Shifts in Zero-Shot Learning: Learning Robust Representations

FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learning

Kernel-Based Tests for Likelihood-Free Hypothesis Testing

Fair GANs through model rebalancing for extremely imbalanced class distributions

Continuous Sweep: an improved, binary quantifier

Novel Class Discovery for Long-tailed Recognition