| 摘 要: 针对现有入侵检测系统(Intrusion Detection System, IDS)检测方法准确率低,泛化能力弱,收敛速度 慢,易陷入局部最优等问题,提出基于PCA(Principal Component Analysis)和多层感知机神经网络(MLP)的入侵检测 模型。该模型首先对数据进行预处理和降维,然后使用该PCA-MLP模型进行训练并使用测试集测试模型的准确率,最 后优化分类器的性能。实验表明,该模型可以提高入侵检测系统的准确率,具有很强的泛化能力。 |
| 关键词: 入侵检测;主成分分析;神经网络;PCA-MLP模型 |
|
中图分类号: TP309
文献标识码: A
|
|
| Research on Intrusion Detection based on Principal Component Analysis and Multilayer Perceptron Neural Network |
|
LIU Hui
|
( Information Of ce, University of Shanghai for Science and Technology, Shanghai 200093, China)
liu_hui@usst.edu.cn
|
| Abstract: Due to the low accuracy, weak generality, slow convergence and easily falling into local optimum of the existing intrusion detection system (IDS), an intrusion detection model based on PCA (Principal Component Analysis) and BP multilayer perceptron neural network is proposed. In this model, data preprocessing and dimension reduction are rstly conducted. Then, the PCA-MLP model is used for training and accuracy is tested. Finally, the performance of the classi er is optimized. The simulation results show that the model can improve the accuracy of intrusion detection system and has strong generality. |
| Keywords: intrusion detection; principal component analysis; neural network; PCA-MLP model |
