«上一篇/Previous Article|本期目录/Table of Contents|下一篇/Next Article»

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

29

期数:

2019年05期

页码:

107-111

栏目:

安全与防范

出版日期:

2019-05-10

文章信息/Info

Title:

Intrusion Detection Technology Based on Self-coded Networks and Clustering

文章编号:

1673-629X(2019)05-0107-05

作者:

周康¹; 万良²

1. 贵州大学 计算机科学与技术学院,贵州 贵阳 550025;2. 贵州大学 软件与理论研究所,贵州 贵阳 550025

Author(s):

ZHOU Kang¹; WAN Liang²

1. School of Computer Science and Technology,Guizhou University,Guiyang 550025,China;2. Institute of Software and Theory,Guiyang 550025,China

关键词:

模糊C均值; 遗传算法; 限制玻尔兹曼机; 自编码网络; 特征降维; 双向映射

Keywords:

fuzzy C-means; genetic algorithm; restricted Boltzmann machine; autoencoder network; feature dimensionality reduction; bidirectional mapping

分类号:

TP301. 6

DOI:

10. 3969 / j. issn. 1673-629X. 2019. 05. 023

摘要:

针对模糊 C 均值聚类算法的入侵检测方法易陷入局部最优,受时间和空间复杂度约束,检测速率低并且使用原始数据集容易陷入“维度灾难冶等问题,提出了一种基于自编码网络(AN)特征降维结合遗传算法(GA)优化模糊 C 均值算法的聚类模型(AN-GA-FCM)。 该模型采用多层限制玻尔兹曼机(RBM)将高维、非线性的数据双向映射到低维空间,建立高维空间到低维空间的自编码网络,进而使用自编码网络权值微调重构低维空间数据的最优高维表示。 并利用遗传算法优化的 FCM 初始聚类中心,避免目标函数陷入局部最优。 将得到的特征降维数据集通过 GA-FCM 进行分类并在KDD爷99 数据集上进行检测,通过与 PCA,SVM,Softmax 等传统算法的实验对比,结果表明,该模型具有较高的入侵检测准确率和较低的分类检测时间。

Abstract:

The intrusion detection method for the fuzzy C-means clustering algorithm is easy to fall into the local optimal,constrained by the time and space complexity,with low detection rate and easy to fall into the “ dimensional disaster ” and other problems using the original data set. For these problems,we propose a novel fuzzy C-means algorithm clustering model (AN-GA-FCM) based on genetic algorithm (GA) optimization combined with auto-encoder network (AN). This model uses multi-layer restricted Boltzmann machine (RBM) to bidirectionally map high-dimensional and nonlinear data into low-dimensional space,establishes high-dimensional space to low-dimensional autoencoder network,and then uses autoencoder network weights to fine-tune parameter,reconstructing the optimal highdimensional representation to low-dimensional spatial data. The FCM initial clustering center optimized by the genetic algorithm is to avoid objective function falling into a local optimum. The dimensionality reduction datasets are classified by GA-FCM detected on the KDD’99 dataset. Meanwhile,compared with the traditional algorithms such as PCA,SVM and Softmax with the model,it shows that the model has higher intrusion detection accuracy and lower classification detection time.

常用功能

更新日期/Last Update: 2019-05-10