基于MCSRELM的网络安全态势预测模型.docVIP

基于MCS-RELM的网络安全态势预测模 型 梁潘冯朝胜 成都航空职业技术学院机电工程学院四川师范大学 摘要： 为了提高网络安全态势的预测准确性，提出一种改进布谷鸟搜索算法和正则极 限学习机相融合的网络安全态势预测模型。首先将一维的网络安全态势数据重构 成多维吋间序列，然后将训练样本集输入到正则极限学习机进行学习，并采用 改进布谷鸟搜索算法优化输入权值和隐含层阈值，建立网络安全态势预测模型， 最后采用Honcynct数据集对模型性能进行测试。结果表明，相对于其它网络安 全态势预测模型，本文模型可更加准确描述网络安全态势的变化趋势，不仅提 高了网络安全态势预测准确性，而且加快了建模速度，可以满足网络安全态势 在线预测要求。 关键词： 网络安全态势;止则极限学习机;布谷鸟搜索算法;预测模型; 作者简介：梁潘（1978-），男，四川广汉人，副教授，硕士，研究方向：计算机 软件与网络安全。E-mail: lpmountain@163. com。 收稿日期:2016-12-10 基金：国家自然科学基金项冃（编 Network Security Situation Prediction Based on Modified Cuckoo Search Algorithm Optimizing Regularized Extreme Learning Machine Liang Pan Feng Chaosheng Department of Mechanical and Electrical Engineering, Chengdu Aeronautic Polytechnic; School of Computer Science, Sichuan Normal University; Abstract： In order to improve the prediction accuracy of net work security situ ation, this paper proposes a novel network security situation prediction model based on modified cuckoo search algor ithm opti mi zing regularized ext reme learning machinc. Firstly, the multidimcnsional time scries of network security situation is obtained based on one-dimensional data, and then the training sample set are input in to the regularized extreme learning machine to learn which modified cuckoo search algorithm is used to optimize input weights and hidden layer threshold and establish the prcdiction model of network security situation finally， by the performance of model is tested by using the Honeynet data. The Simulation results show that, compared with other network security situation prediction models, the proposed model can more accurately describe the change trend of network security situation, improve the prediction accuracy of network security situation， emd fastens the modeling speed of network security situation, can meet the online prediction requirements of network security situation. Keyword： network security situation; regulerri7ed extreme learning machine; Cuckoo search algorithm; prcdiction model; Received： 2016-12