基于粒子群优化随机森林的恶意软件检测分析

Malware Detection and Analysis Based on Particle Swarm Optimization Random Forest

为了有效地检测软件家族中的恶意软件,改进了加权随机森林模型,提出基于粒子群优化的随机森林(particle swarm optimization-random forest,PSO-RF)模型,并使用基于粒子群优化随机森林的恶意软件检测方法对恶意软件家族进行分类。对得出的结果与决策树、支持向量机等经典分类器从准确率、精确度、召回率、综合评价指标值(F;值)等指标进行对比分析,以验证改进后的算法的有效性与合理性。结果表明,PSO-RF模型评估指标均是最高的,能大大提升恶意软件的检测效果。

In order to effectively detect malicious software in software families,a weighted random forest model is improved,a random forest(PSO-RF)model based on particle swarm optimization is proposed,and malicious software families are classified using a malicious software detection method based on particle swarm optimization random forest.The results are compared with the classical classifiers such as decision tree and support vector machine in terms of accuracy,precision,recall and comprehensive evaluation index(F;-measure),so as to verify the effectiveness and rationality of the improved algorithm.The results show that the evaluation index of PSO-RF model is the highest,which can greatly improve the detection effect of malware.