Authors: Catherine Kuenz Murphy
Addresses: Graduate School of Management and Technology, University of Maryland University College, Adelphi, MD 20783, USA
Abstract: Accounting standards require auditors to attach an explanatory paragraph to their audit report if they have substantial doubt about an entity|s ability to continue as a going-concern. In the paragraph, auditors provide the rationale for the opinion. Given the corporation|s financial ratios, machine learning can quantify this rationale. This study induces decision rules from cases of financially distressed companies for the period of economic expansion, in the USA, from 1991 to 2000. The induced rules for the opinion agree with two factors frequently cited in the corresponding audit reports: stockholders| deficiency and default. The values of only two rules can correctly identify 80% of the decisions. The induced trees yield specific rules corresponding to an expert system for the going-concern decision.
Keywords: machine learning; knowledge acquisition; induced decision trees; expert systems; auditor opinions; going concerns; auditing; accounting standards; USA; United States; decision rules; financially distressed companies.
International Journal of Computer Applications in Technology, 2008 Vol.33 No.2/3, pp.138 - 145
Published online: 10 Dec 2008 *Full-text access for editors Access for subscribers Purchase this article Comment on this article