Detecting labeling bias using influence functions
Labeling bias arises during data collection due to resource limitations or unconscious bias, leading to unequal label error rates across subgroups or misrepresentation of subgroup prevalence. Most fairness constraints assume training labels reflect the true distribution, rendering them ineffective when labeling bias is present; leaving a challenging question, that \textit{how can we detect such labeling bias?} In this work, we investigate whether influence functions can be used to detect labeling bias. Influence functions estimate how much each training sample affects a model's predictions by leveraging the gradient and Hessian of the loss function -- when labeling errors occur, influence functions can identify wrongly labeled samples in the training set, revealing the underlying failure mode. We develop a sample valuation pipeline and test it first on the MNIST dataset, then scaled to the more complex CheXpert medical imaging dataset. To examine label noise, we introduced controlled errors by flipping 20\% of the labels for one class in the dataset. Using a diagonal Hessian approximation, we demonstrated promising results, successfully detecting nearly 90\% of mislabeled samples in MNIST. On CheXpert, mislabeled samples consistently exhibit higher influence scores. These results highlight the potential of influence functions for identifying label errors.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Resolving Training Biases via Influence-based Data Relabeling
The performance of supervised learning methods easily suffers from the training bias issue caused by train-test distribution mismatch or label noise. Influence function is a technique that estimates the impacts of a tra…
Understanding Impact of Human Feedback via Influence Functions
In Reinforcement Learning from Human Feedback (RLHF), it is crucial to learn suitable reward models from human feedback to align large language models (LLMs) with human intentions. However, human feedback can often be no…
Developing Linguistic Patterns to Mitigate Inherent Human Bias in Offensive Language Detection
With the proliferation of social media, there has been a sharp increase in offensive content, particularly targeting vulnerable groups, exacerbating social problems such as hatred, racism, and sexism. Detecting offensive…
Data AugmentationFairnessInfluence Functions for Preference Dataset Pruning
Language models are commonly fine-tuned via reinforcement learning to alter their behavior or elicit new capabilities. Datasets used for these purposes, and particularly human preference datasets, are often noisy. The re…
parameter-efficient fine-tuningReinforcement LearningFast Model Debias with Machine Unlearning
Recent discoveries have revealed that deep neural networks might behave in a biased manner in many real-world scenarios. For instance, deep networks trained on a large-scale face recognition dataset CelebA tend to predic…
counterfactualDecision MakingFace RecognitionMachine Unlearning+1