Automatic identification of enterprise operation risk based on data analysis

摘要

The enterprise operational risk identification is the enterprise financing a key issue in the study of risk measurement and monitoring, in order to solve this problem, the paper on the study of the enterprise operational risk identification process with Logistic optimal regression extract significantly explain index, use K - Means cluster analysis to chosen for cluster analysis, the abnormal situation of enterprise risk group classification, analysis of the characteristicsof the index data within a group, according to explain the index variable data classification, so to conform to the index dataclassification discriminant operational risk of enterprise can normal. For enterprise operational risk classification is difficult to divide, further using the decision tree model and the CART model this group of enterprises were analyzed, and compared the model classification results of two methods of analysis, it is concluded that enterprise operational risk identification results in the end.