The application of fuzzy decision tree analysis in an exposition of the antecedents of audit fees

Since the seminal work of Zadeh fuzzy set theory (FST) has evolved into a valuable extension to traditional techniques, such as regression and decision tree models, for decision analysis conducted under conditions of vagueness and ambiguity. This paper is concerned with the exposition and application of a fuzzy decision tree approach to a problem involving typical accounting data. More specifically, a set of fuzzy ‘if…then…’ rules is constructed to classify the level of corporate audit costs based on a number of characteristics of the companies and their auditors. The fuzzy rules enable a decision-maker to gain additional insights into the relationship between firm characteristics and audit fees, through human subjective judgements expressed in linguistic terms. We also extend previous research by developing a more objective semi-automated method of constructing the FST related membership functions which mitigates reliance on the input of human expert opinions.