Class Balancing

Class imbalance, where some classes in a dataset are significantly under-represented compared to others, is a major challenge in machine learning, hindering model performance and fairness. Current research focuses on developing techniques to mitigate this imbalance, exploring methods such as data augmentation (e.g., using GANs or mosaic augmentation), adaptive weighting schemes, and topological augmentation of graph-based models, all aiming to improve model accuracy and reduce bias without relying solely on class rebalancing. These advancements are crucial for improving the reliability and applicability of machine learning models across diverse domains, particularly in areas like medical imaging and agriculture where data scarcity and class imbalance are common.

Papers

July 22, 2024

Beyond Size and Class Balance: Alpha as a New Dataset Quality Metric for Deep Learning
Josiah Couch, Rima Arnaout, Ramy Arnaout
Deep Learning Training Data Size Matter Alpha MEPOL Dataset Diversity Diversity Score Class Balancing Dataset Quality

June 18, 2024

Agriculture-Vision Challenge 2024 -- The Runner-Up Solution for Agricultural Pattern Recognition via Class Balancing and Model Ensemble
Wang Liu, Zhiyu Wang, Puhong Duan, Xudong Kang, Shutao Li
Semantic Segmentation Semantic Segmentation Model Place Solution Model Ensemble Agricultural Image Class Balancing Mosaic Data Augmentation

August 27, 2023

Class-Imbalanced Graph Learning without Class Rebalancing
Zhining Liu, Ruizhong Qiu, Zhichen Zeng, Hyunsik Yoo, David Zhou, Zhe Xu, Yada Zhu, Kommy Weldemariam, Jingrui He, Hanghang Tong
Class Imbalance Graph Learning Model Graph Topology Imbalanced Graph Class Balancing

June 24, 2023

Evaluating the Utility of GAN Generated Synthetic Tabular Data for Class Balancing and Low Resource Settings
Nagarjuna Chereddy, Bharath Kumar Bolla
Training Data GAN Model Class Imbalance Task Utility GAN Based Synthetic Tabular Data Class Balancing Class Balancing Technique

January 22, 2023

Deterministic Online Classification: Non-iteratively Reweighted Recursive Least-Squares for Binary Class Rebalancing
Se-In Jang
Least Square Boolean Classifier Class Balancing Deterministic Online

November 30, 2022

Overlapping oriented imbalanced ensemble learning method based on projective clustering and stagewise hybrid sampling
Fan Li, Bo Wang, Pin Wang, Yongming Li
Ensemble Learning Class Imbalance Correlation Clustering Class Balancing Projective Clustering

September 18, 2022

Adaptive Dimension Reduction and Variational Inference for Transductive Few-Shot Classification
Yuqing Hu, Stéphane Pateux, Vincent Gripon
Classification Code Variational Inference Dimension Reduction Feature Distribution Variational Bayesian Inference Probabilistic Linear Discriminant Analysis Transductive Approach Class Balancing

February 3, 2022

GALAXY: Graph-based Active Learning at the Extreme
Jifan Zhang, Julian Katz-Samuels, Robert Nowak
Active Learning Distant GALAXY Balanced Dataset Extreme Value Graph Based Active Learning Class Balancing

Class Balancing

Papers

Beyond Size and Class Balance: Alpha as a New Dataset Quality Metric for Deep Learning

Agriculture-Vision Challenge 2024 -- The Runner-Up Solution for Agricultural Pattern Recognition via Class Balancing and Model Ensemble

Class-Imbalanced Graph Learning without Class Rebalancing

Evaluating the Utility of GAN Generated Synthetic Tabular Data for Class Balancing and Low Resource Settings

Deterministic Online Classification: Non-iteratively Reweighted Recursive Least-Squares for Binary Class Rebalancing

Overlapping oriented imbalanced ensemble learning method based on projective clustering and stagewise hybrid sampling

Adaptive Dimension Reduction and Variational Inference for Transductive Few-Shot Classification

GALAXY: Graph-based Active Learning at the Extreme