English

Neural Relation Graph: A Unified Framework for Identifying Label Noise and Outlier Data

Machine Learning 2023-10-31 v5

Abstract

Diagnosing and cleaning data is a crucial step for building robust machine learning systems. However, identifying problems within large-scale datasets with real-world distributions is challenging due to the presence of complex issues such as label errors, under-representation, and outliers. In this paper, we propose a unified approach for identifying the problematic data by utilizing a largely ignored source of information: a relational structure of data in the feature-embedded space. To this end, we present scalable and effective algorithms for detecting label errors and outlier data based on the relational graph structure of data. We further introduce a visualization tool that provides contextual information of a data point in the feature-embedded space, serving as an effective tool for interactively diagnosing data. We evaluate the label error and outlier/out-of-distribution (OOD) detection performances of our approach on the large-scale image, speech, and language domain tasks, including ImageNet, ESC-50, and SST2. Our approach achieves state-of-the-art detection performance on all tasks considered and demonstrates its effectiveness in debugging large-scale real-world datasets across various domains. We release codes at https://github.com/snu-mllab/Neural-Relation-Graph.

Keywords

Cite

@article{arxiv.2301.12321,
  title  = {Neural Relation Graph: A Unified Framework for Identifying Label Noise and Outlier Data},
  author = {Jang-Hyun Kim and Sangdoo Yun and Hyun Oh Song},
  journal= {arXiv preprint arXiv:2301.12321},
  year   = {2023}
}

Comments

NeurIPS 2023

R2 v1 2026-06-28T08:25:01.572Z