人工蜂群联合入侵杂草优化的云平台异常行为数据挖掘

Cloud platform abnormal behavior data mining based on artificial bee colony joint invasive weed optimization

云计算平台异常行为数据的维度增高,会影响数据挖掘效果。为此,提出人工蜂群联合入侵杂草优化的云平台异常行为数据挖掘方法。构造多标签核映射数据降维方法,并将径向基函数作为核函数,对数据降维;采用混合蜂群杂草算法对径向基函数带宽和最小二乘多分类孪生支持向量机惩罚因子进行优化;采用最优径向基函数带宽优化多标签核映射数据降维算法,并利用该算法对数据进行降维,将其输入到优化后的最小二乘多分类孪生支持向量机决策函数中,计算数据与各个超平面之间的距离,确定数据所属类别,从而获取最优的云计算平台异常行为数据挖掘结果。实验结果表明,该方法在挖掘误差、能量损耗、挖掘时间等指标上效果较好。

The dimension of abnormal behavior data of cloud computing platform increases,which can affect the the effectiveness of data mining.A method of cloud platform abnormal behavior data mining based on artificial bee colony joint invasive weed optimization is proposed.The data dimensionality reduction method of multilabel kernel mapping is constructed,and the radial basis function is used as the kernel function to reduce the data dimensionality.The hybrid bee colony weed algorithm is used to optimize the radial basis function bandwidth and the penalty factor of the least squares multi classification twin support vector machine.The optimal radial basis function bandwidth is used to optimize the multi label kernel mapping data dimensionality reduction algorithm,and the algorithm is used to reduce the data dimensionality.It is input into the optimized least squares multi classification twin support vector machine decision function to calculate the distance between the data and each hyperplane,and determine the category of the data,so as to obtain the optimal data mining results of cloud computing platform abnormal behavior.The experimental results show that the proposed method can perform well in indicators such as mining error,energy loss and mining time.