Detecting Accounting Fraud in China A-Share Market With PU Learning

Zhaolong Zhang; Zhenyu Liu; Fengmin Xu; Xiangyu Chang

doi:10.1111/acfi.70045

Detecting Accounting Fraud in China A-Share Market With PU Learning

Zhaolong Zhang
, Zhenyu Liu
, Fengmin Xu
, Xiangyu Chang

School of Management

Xi'an Jiaotong University

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

Detecting accounting fraud is critical for financial market integrity. Traditional methods struggle due to imbalanced data and incomplete labelling. This study introduces positive and unlabelled (PU) learning to enhance detection accuracy using labelled fraud cases and extensive unlabelled samples. Analysing China's A-share market data from 2001 to 2022, segmented into stability (2001–2019) and transition (2020–2022) regulatory periods, results confirm that PU learning significantly improves the detection of fraudulent firms, demonstrating robust performance under varying regulatory and economic conditions. The findings highlight PU learning's effectiveness as a robust method for detecting fraud, offering practical insights for enhancing regulatory oversight and maintaining market stability.

Original language	English
Pages (from-to)	3361-3378
Number of pages	18
Journal	Accounting and Finance
Volume	65
Issue number	4
DOIs	https://doi.org/10.1111/acfi.70045
State	Published - Dec 2025

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 8 Decent Work and Economic Growth
SDG 12 Responsible Consumption and Production

Keywords

China A-share market
PU learning
accounting fraud detection
enhanced recall
resource efficiency

Access to Document

10.1111/acfi.70045

Cite this

@article{ec9e944267d34c64a04ebb0b56ce6166,

title = "Detecting Accounting Fraud in China A-Share Market With PU Learning",

abstract = "Detecting accounting fraud is critical for financial market integrity. Traditional methods struggle due to imbalanced data and incomplete labelling. This study introduces positive and unlabelled (PU) learning to enhance detection accuracy using labelled fraud cases and extensive unlabelled samples. Analysing China's A-share market data from 2001 to 2022, segmented into stability (2001–2019) and transition (2020–2022) regulatory periods, results confirm that PU learning significantly improves the detection of fraudulent firms, demonstrating robust performance under varying regulatory and economic conditions. The findings highlight PU learning's effectiveness as a robust method for detecting fraud, offering practical insights for enhancing regulatory oversight and maintaining market stability.",

keywords = "China A-share market, PU learning, accounting fraud detection, enhanced recall, resource efficiency",

author = "Zhaolong Zhang and Zhenyu Liu and Fengmin Xu and Xiangyu Chang",

note = "Publisher Copyright: {\textcopyright} 2025 Accounting and Finance Association of Australia and New Zealand.",

year = "2025",

month = dec,

doi = "10.1111/acfi.70045",

language = "英语",

volume = "65",

pages = "3361--3378",

journal = "Accounting and Finance",

issn = "0810-5391",

publisher = "John Wiley and Sons Inc",

number = "4",

}

TY - JOUR

T1 - Detecting Accounting Fraud in China A-Share Market With PU Learning

AU - Zhang, Zhaolong

AU - Liu, Zhenyu

AU - Xu, Fengmin

AU - Chang, Xiangyu

PY - 2025/12

Y1 - 2025/12

N2 - Detecting accounting fraud is critical for financial market integrity. Traditional methods struggle due to imbalanced data and incomplete labelling. This study introduces positive and unlabelled (PU) learning to enhance detection accuracy using labelled fraud cases and extensive unlabelled samples. Analysing China's A-share market data from 2001 to 2022, segmented into stability (2001–2019) and transition (2020–2022) regulatory periods, results confirm that PU learning significantly improves the detection of fraudulent firms, demonstrating robust performance under varying regulatory and economic conditions. The findings highlight PU learning's effectiveness as a robust method for detecting fraud, offering practical insights for enhancing regulatory oversight and maintaining market stability.

AB - Detecting accounting fraud is critical for financial market integrity. Traditional methods struggle due to imbalanced data and incomplete labelling. This study introduces positive and unlabelled (PU) learning to enhance detection accuracy using labelled fraud cases and extensive unlabelled samples. Analysing China's A-share market data from 2001 to 2022, segmented into stability (2001–2019) and transition (2020–2022) regulatory periods, results confirm that PU learning significantly improves the detection of fraudulent firms, demonstrating robust performance under varying regulatory and economic conditions. The findings highlight PU learning's effectiveness as a robust method for detecting fraud, offering practical insights for enhancing regulatory oversight and maintaining market stability.

KW - China A-share market

KW - PU learning

KW - accounting fraud detection

KW - enhanced recall

KW - resource efficiency

UR - https://www.scopus.com/pages/publications/105008437579

U2 - 10.1111/acfi.70045

DO - 10.1111/acfi.70045

M3 - 文章

AN - SCOPUS:105008437579

SN - 0810-5391

VL - 65

SP - 3361

EP - 3378

JO - Accounting and Finance

JF - Accounting and Finance

IS - 4

ER -

Detecting Accounting Fraud in China A-Share Market With PU Learning

Abstract

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this