From the end of 2007 the U.S. sub-prime mortgage crisis triggered a first wave of financial tsunami until now; this had a significant impact on financial economy. Correctly analyzing and forecasting the financial crisis is more important. The investigation provides two instructions for researchers and executives. The investors adopted RELIEF method and support vector machine as an early warning system. Sequentially, rough set theory is used to identify informative information and decision rules for managers to improve the financial situations. The results showed that hybrid RELIEF and Support Vector Machine perform a satisfactory job in detecting financial crisis; the decision rules derive from rough set theory can provide instruction for managers to change their management policy.