Empirical Study Of The Financial Report Information Manipulating Early-warning On Listed Companies

Base on the current research, this article defined the concept of the financial report information manipulation, distinguished it form other concepts, cleared who are the objects that the model serve to primarily and elaborated the foundation and assumptions of it. The concept of information group which was put forward creatively was used to solve the problem of identifying indicators signal. The relative fluctuation rate formula was designed to index analysis of the quantifiable-financial-indicators information group. The financial report information manipulating early-warning signal system which was built with the idea of Balanced Scorecard, was stratified the indicators gradually, and was identified the specific early-warning indicators of discriminating according to the track that happened on the financial report information manipulation. Artificial Neural Networks and Support Vector Machine were selected to test classification accuracy and the test results were analyzed.