基于小波核主成分分析和差分进化优化极限学习机的入侵检测

INTRUSION DETECTION BASED ON WAVELET KERNEL PCA AND DE OPTIMISED EXTREME LEARNING MACHINE

针对网络入侵检测,提出一种基于小波核主成分分析和差分进化极限学习机相结合的方法。首先采用核主成分分析法对原始数据进行非线性降维处理,为了进一步提高核PCA的非线性映射能力,引用小波核函数作为核PCA的核函数。然后采用极限学习机对处理后的数据进行分类识别,针对初始权值随机选择造成极限学习机性能不稳定的问题,采用差分进化算法来获得最优的初始权值。实验结果表明该算法可以有效提高入侵检测的识别率,降低误报率和漏报率。

For network intrusion detection,we propose such a method which combines the wavelet kernel PCA and DE optimised extreme learning machine. First,the kernel principal component analysis（ PCA） is applied to conduct the nonlinear dimensionality reduction on original data,in order to further improve nonlinear mapping ability of kernel PCA,wavelet kernel function is introduced as its kernel function. Then the extreme learning machine is used for the classification and recognition of the processed data,and the differential evolution（ DE） algorithm is used to obtain the optimal initial weights for the unstable performance of the extreme learning machine caused by random selection of initial weights. Experimental results show that the algorithm proposed can effectively improve the recognition rate of intrusion detection and reduce the rates of false positives and false negatives.