|
Providing an Audit Quality Model for Use in Investors' Decisions Using Statistical Models and Machine Learning
|
Mahnaz Norouzi Aslbalkanlou1   , Parviz Piri2 , Ali Ashtab3 , Pari Chalaki4 |
1- Ph.D. Candidate. Department of Accounting, Faculty Economic and Management, Urmia University, Urmia, Iran
2- Associate Prof, Deparment of Accounting, Faculty of Economic and Management, Urmia University, Urmia, Iran , p.piri@urmia.ac.ir
3- Assistant Professor
4- Assistant Prof, Deparment of Accounting, Faculty of Economic and Management, Urmia University, Urmia, Iran |
|
|
Abstract: (102 Views) |
Audit quality is important for users of accounting information to evaluate performance, predict profitability and determine the true value of the company. With this statement, the main goal of the current research is to present the audit quality model (input, process, output and background factors) to be used in the decisions of investors and financial market participants. To reach the goal of the research, a variety of statistical models and machine learning have been used to achieve an optimal model in predicting the investment decision model. In order to evaluate the performance of machine learning models, two criteria of model prediction accuracy and area under the curve have been used. Finally, in order to choose the model that has the best performance for predicting the investment decision model, the system performance characteristic curve has been used. The results of the research showed that after calculating the average prediction accuracy, the models of inference rules, K nearest neighbor and evolutionary support vector machine have the highest prediction accuracy of 84..29%, 78.74% and 77.01% among machine learning models, respectively. Also, based on the results of the system performance characteristic curve, the inferential rules model has the best performance in predicting the investment decision model with a prediction accuracy of 84.29% and was selected as the optimal model. One of the ways to help investment analysts and financial market participants is to provide predictive models about the company's information outlook. The closer the predictions are to reality, the more correct decisions will be made. The high quality of auditing can lead to the consolidation of transparent financial reporting and increase the accuracy in evaluating the financial status of companies, which, in turn, affects the quality of financial decisions of investors and increasing the efficiency of financial markets. Statistical methods and data mining can largely provide a support system for investors' decision making. Therefore, in this research, different types of statistical and machine learning models were developed. The results of research can provide the reader with a better understanding of the effect of audit quality from the perspective of input, process, output and background factors on the decision model according to the information content and decision usefulness theory for decision making.
|
Article number: 10 |
| Keywords: Audit Quality, Statistical Model, Decision Model, Machine Learning |
|
|
Full-Text [PDF 804 kb]
(33 Downloads)
|
Article type: Research |
Subject:
Special
Received: 2024/10/2 | Accepted: 2024/12/9 | Published: 2025/05/26
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
