Corporate failure prediction of Chinese listed companies: A variable precision rough set theory

Since the seminal work of Pawlak has been published in 1982, the rough set theory (RST) has continued to flourish as a tool for data mining, however, to date, relatively a few empirical researches have been conducted on the rough set approach in the context of corporate failure prediction in Chinese market. This paper applies an advanced RST, namely the variable precision rough sets (VPRS) model, to predict between failed and non-failed Chinese listed companies. In addition to the applying of the VPRS model, we utilize the FUSINTER method to discretize the data we collected from China Center for Economics Research (CCER) database. Our research explores how financial and non-financial indicators impact on the corporate performance and concludes that the VPRS is a practical and promising method in corporate failure predictions.