Long-tail learning via logit adjustment
Real-world classification problems typically exhibit an imbalanced or long-tailed label
distribution, wherein many labels are associated with only a few samples. This poses a …
distribution, wherein many labels are associated with only a few samples. This poses a …
Certifying and removing disparate impact
M Feldman, SA Friedler, J Moeller… - proceedings of the 21th …, 2015 - dl.acm.org
What does it mean for an algorithm to be biased? In US law, unintentional bias is encoded
via disparate impact, which occurs when a selection process has widely different outcomes …
via disparate impact, which occurs when a selection process has widely different outcomes …
Optimal classifier for imbalanced data using Matthews Correlation Coefficient metric
Data imbalance is frequently encountered in biomedical applications. Resampling
techniques can be used in binary classification to tackle this issue. However such solutions …
techniques can be used in binary classification to tackle this issue. However such solutions …
Contrastive learning based hybrid networks for long-tailed image classification
Learning discriminative image representations plays a vital role in long-tailed image
classification because it can ease the classifier learning in imbalanced cases. Given the …
classification because it can ease the classifier learning in imbalanced cases. Given the …
A unified generalization analysis of re-weighting and logit-adjustment for imbalanced learning
Real-world datasets are typically imbalanced in the sense that only a few classes have
numerous samples, while many classes are associated with only a few samples. As a result …
numerous samples, while many classes are associated with only a few samples. As a result …
Autobalance: Optimized loss functions for imbalanced data
M Li, X Zhang, C Thrampoulidis… - Advances in Neural …, 2021 - proceedings.neurips.cc
Imbalanced datasets are commonplace in modern machine learning problems. The
presence of under-represented classes or groups with sensitive attributes results in …
presence of under-represented classes or groups with sensitive attributes results in …
Learning from corrupted binary labels via class-probability estimation
Many supervised learning problems involve learning from samples whose labels are
corrupted in some way. For example, each sample may have some constant probability of …
corrupted in some way. For example, each sample may have some constant probability of …
Fairness risk measures
R Williamson, A Menon - International conference on …, 2019 - proceedings.mlr.press
Ensuring that classifiers are non-discriminatory or fair with respect to a sensitive feature (eg,
race or gender) is a topical problem. Progress in this task requires fixing a definition of …
race or gender) is a topical problem. Progress in this task requires fixing a definition of …
Ods: Test-time adaptation in the presence of open-world data shift
Test-time adaptation (TTA) adapts a source model to the distribution shift in testing data
without using any source data. There have been plenty of algorithms concentrated on …
without using any source data. There have been plenty of algorithms concentrated on …
Consistent binary classification with generalized performance metrics
OO Koyejo, N Natarajan… - Advances in neural …, 2014 - proceedings.neurips.cc
Performance metrics for binary classification are designed to capture tradeoffs between four
fundamental population quantities: true positives, false positives, true negatives and false …
fundamental population quantities: true positives, false positives, true negatives and false …