Algorithm Adaptation Methods for Prediction of Opinion and Exceptions in Financial Statements Audit Reports: Case for Financial Statements Audit in Riau Islands Province

Prior research on financial statement audits has mostly employed multi-label classification, like the Problem Transformation Methods, and single-label classification, such as opinion prediction, identification, and detection. Using information from financial statement audit reports in the Riau Islands province, we suggest applying multi-label classification to predict 'opinion and exceptions.' We employ opinion and exceptions as labels and financial ratios as attributes. Back-Propagation Neural Network (BPNN) and Deep Back-Propagation Neural Network (DBPNN), two Algorithm Adaptation Methods, are employed in this study. Two Algorithm Adaptation Methods are computed using two model architectures, Multilayer Perceptron (MLP) and Voted Perceptron (VP). For Hamming Loss (HL), One Error (OE), Rank Loss (RL), and Average Precision (AP), the best evaluation metric outcomes are 0.0417, 0.0000, 0.0333, and 0.9786, respectively.