数据挖掘算法在大数据网络安全防御中的应用研究

The Application of Data Mining Algorithms in Big Data Network Security Defense

为了研究数据挖掘算法在大数据网络安全防御中的应用,研究利用随机森林算法进行数据分类和识别,并利用果蝇优化算法对随机森林算法进行改进。通过提取关键特征,对数据进行预处理等操作,以提升改进随机森林分类算法在网络入侵检测中的有效性。研究结果表明,和支持向量机分类器、传统的MLPClassifier分类器相比,改进随机森林算法能够更为准确地识别网络入侵类型,更适合应用于网络入侵检测和网络防御。仿真结果表明,改进随机森林算法具有明显的优势,在不同攻击类型中显示出较高的精度。改进随机森林算法对不同攻击连续的检测精度保持在0.79以上,传统的MLPClassifier分类器对不同攻击连续的检测精度保持在0.66以上,支持向量机分类器不同攻击类型的检测精度保持在0.72以上。表明了改进随机森林算法强大的网络入侵检测能力,并能够为网络安全防御提供重要保障。

In order to study the application of data mining algorithm in big data network security defense,the random forest algorithm was used for data classification and identification,and the random forest algorithm was improved by using the fruit fly optimization algorithm.By extracting key features and preprocessing data,the effectiveness of the improved random forest classification algorithm in network intrusion detection can be improved.The research results indicate that compared with support vector machine classifiers and traditional MLPC classifier,the improved random forest algorithm can more accurately identify network intrusion types and it more suitable for application in network intrusion detection and defense.The simulation results show that the improved random forest algorithm has significant advantages and shows high accuracy in different attack types.The improved random forest algorithm maintains a detection accuracy of over 0.79 for different attack sequences,while the traditional MLPC classifier maintains a detection accuracy of over 0.66 for different attack sequences,and the support vector machine classifier maintains a detection accuracy of over 0.72 for different attack types.This demonstrates the powerful cyber penetration detection ability of the improved random forest algorithm,which can provide important guarantee for network security defense.