模因组融合信息度传递的溢出类漏洞线性预测

Linear Prediction of Overflow Vulnerability Based on Meme Group Information Fusion Transmission

对缓冲区的溢出类漏洞的线性预测是提高漏洞挖掘效率的重要一环,传统方法采用粒子群扰动聚类方法进行溢出类漏洞的预测和挖掘,存在预测精度不准的问题。提出一中基于模因组融合信息度传递的缓冲区溢出漏洞线性预测方法,实现对安全漏洞的准确检测。采用四叉树算法对混合粒子群多维数据进行数据预处理,采用模因组融合信息度传递,结合高斯变异对其进行扰动以代替随机产生新粒子个体的操作,实现对海量多模态数据的优化聚类和线性预测。实验结果表明,算法能准确跟踪溢出类漏洞的演化轨迹,实现对溢出类漏洞的线性预测,预测精度提高24.3%,漏洞挖掘性能提高,保证了应用信息应用环境安全。

Linear prediction of overflow vulnerability of the buffer zone is an important step for increasing vulnerability min-ing efficiency, and traditional method uses particle swarm clustering method for mining disturbance overflow holes, the pre-diction accuracy is bad. An improved linear prediction of overflow vulnerability method is proposed based on meme groupinformation fusion transmission, so accurate detection of security vulnerabilities is realized. The 4 tree algorithm is used fordata pretreatment on hybrid particle swarm of multidimensional data, the meme group fusion information transfer, combinedwith Gauss mutation to replace the random disturbance to generate new particles on the individual operation, realize the op-timization of clustering and linear prediction of massive multi modal test data. The experimental results show that the algo-rithm can accurately track the evolution, linear trajectory prediction is achieved, the prediction accuracy is improved by24.3%, vulnerability mining performance is improved, so the information application environment security is guaranteed.