A neural network approach to the prediction of going concern status

摘要: Abstract The assessment of a firm's going concern status is not an easy task. To assist auditors, prediction models based on statistical methods such as multiple discriminant analysis and logit/probit have been explored with some success. This study attempts to look at different more recent approach—neural networks. In particular, neural network model the feedforward, backpropagation type was constructed predict from six financial ratios, using data set containing 165 non-going concerns matched concerns. On evenly distributed hold-out sample, trained correctly predicted all 30 test cases. results suggest that networks can be promising avenue research application in area.