Improvement And Application In Intrusion Detection Of Data Mining Algorithm

Data mining is playing a more and more important role in the decision-making activities of every walk of life. Data mining automatically searches for predictive information among mass of data to predict the tendencies and actions in the future, thereby, it can support our decision-making. Facing rapidly expanding information, data mining algorithm will be the bottle-neck. Therefore, it is very useful for us to improve the data mining algorithm, especially when related to the intrusion detection in the rapid network environment nowadays, improved algorithm can offer higher data package analyzing ability and higher system performance. Meanwhile, in the information age, intrusion detection system is a kind of active safety technique, and it provides protection from internal attacks, external attacks and mistaken operation. Therefore, it can offer service from different angles, which makes it the research focus of security field.This paper focuses on the association rules mining algorithm of data mining, and intrusion detection technology. It describes the process of data mining and technique, analyzes the deficiency of the algorithm, brings forward an association rules mining algorithm based on frequent itemset mining through matrix row and column vector according to the theory of association rules mining and knowledge of matrix and vector.Meanwhile, this paper evaluates the algorithm through experiment. In the end, according to the Common Intrusion Detection Framework, the association rules mining technique is applied to the intrusion detection system, and then an intrusion detection system based on data mining is designed.