Study on Rule Extraction Based on Rough Set in the Risk Management

To deal with the risk recognition from the complex risk management, a novel risk recognition method was studied through rule extration based on rough set. According to the indiscernible relation in rough set, a formula to compute attribute importance is put forward based on the concept of similar matrix and attribute frequency in the matrix, which is used as heuristic knowledge in deciding the sequence of attribute reduction. The idea of half-in query is used in the algorithm to accurate reduction. Attribute value reduction was realized through gradually deleting the redundant attribute value for every rule in the information table by the correlation of condition attributes and decision attributes. Finally, a concise rule set for risk recognition was obtained. The illustration and experiment results indicate that the method is effective and efficient.