Abstract
The traditional single minimum support data mining algorithm has some problems, such as too much space occupied by data, resulting in insufficient accuracy of the algorithm, which is difficult to meet the needs of the development of the times. Therefore, an intrusion data mining algorithm based on multiple minimum support is proposed. First, the feature parameters of frequent itemsets of intrusion data are extracted, and the sequence itemsets are divided according to the feature parameters. Then, the data mining features are transformed with the equivalent binary data transformation method, and the multi-support tree structure is optimized according to the data processing results. Data classification mining is carried out with the data tree structure information, and the intrusion data features are deeply mined. Finally, the research of the intrusion data mining algorithm based on the multi-minimum support is completed. Through comparative experiments, it is proved that the accuracy of the intrusion data mining algorithm based on multiple minimum support is 35 % - 75 % higher than that of the traditional single minimum support data mining algorithm.
Modules
Algorithms
Feature extraction , Clustering algorithms ,Intrusion detection
Software And Hardware
python,anaconda,jupyter,mysql,pandas