Audit Opinion Prediction Using the Decision Tree Algorithm
Audit Opinion Prediction Using the Decision Tree Algorithm

Author(s): Amra Gadžo, Mirza Suljić, Erna Heric
Subject(s): Social Sciences
Published by: Udruženje ekonomista i menadžera Balkana
Keywords: Audit opinion; Decision tree; Prediction
Summary/Abstract: This paper presents a data mining approach for Audit opinion prediction in Government-owned enterprises within the Federation of Bosnia and Herzegovina using the Decision tree algorithm. A database was constructed from financial statements covering 2004-2019, incorporating indicators from balance sheets, income statements, and cash flow statements, alongside corresponding Audit opinions from the state audit body. The study evaluates three Decision tree algorithms (J48, RandomTree, REPTree) on data from 2020-2023, with REPTree achieving 73% classification accuracy through seven predictive rules. The findings demonstrate the potential of data mining techniques for pattern recognition in audit reports, contributing to transparency in financial reporting and supporting regulatory authorities in detecting irregularities within Government-owned enterprises.

• Details
• Contents