CHG Shapley: Efficient Data Valuation and Selection towards Trustworthy Machine Learning

2024 article cai2024chgshapley ○ Not yet verified

Authors

Huaiguang Cai

Venue

arXiv preprint

Abstract

Proposes CHG (compound of Hardness and Gradient) utility function to approximate the utility of each data subset, reducing computational complexity to a single model retraining—achieving a quadratic improvement over existing Data Shapley methods.

BibTeX

Local Entry

@article{cai2024chgshapley,
  title = {CHG Shapley: Efficient Data Valuation and Selection towards Trustworthy Machine Learning},
  author = {Huaiguang Cai},
  year = {2024},
  journal = {arXiv preprint},
  url = {https://arxiv.org/abs/2406.11730},
  abstract = {Proposes CHG (compound of Hardness and Gradient) utility function to approximate the utility of each data subset, reducing computational complexity to a single model retraining—achieving a quadratic improvement over existing Data Shapley methods.}
}

From AUTO:S2

@article{cai2024chgshapley,
  title = {CHG Shapley: Efficient Data Valuation and Selection towards Trustworthy Machine Learning},
  author = {Huaiguang Cai},
  year = {2024},
  journal = {arXiv.org},
  doi = {10.48550/arXiv.2406.11730}
}