首页|期刊简介|投稿指南|分类索引|刊文选读|订阅指南|证明资料|样刊邮寄查询|常见问题解答|联系我们
范志鹏,李军,刘宇强,等. 基于灰度纹理指纹的恶意代码分类[J]. 科学技术与工程, 2020, 20(29): 12014-12020.
Fan Zhipeng,李军,Liu Yuqiang,et al.Classification of Malware Based on Gray Texture Fingerprint[J].Science Technology and Engineering,2020,20(29):12014-12020.
基于灰度纹理指纹的恶意代码分类
Classification of Malware Based on Gray Texture Fingerprint
投稿时间:2020-01-03  修订日期:2020-06-24
DOI:
中文关键词:  恶意代码  灰度共生矩阵  神经网络  纹理指纹
英文关键词:malware  gray level co-occurrence matrix  neural networks  texture fingerprint
基金项目:国家自然科学基金(61902116);国家自然科学基金项目(51508169)
           
作者单位
范志鹏 湖北工业大学
李军 湖北工业大学
刘宇强 湖北工业大学
钮焱 湖北工业大学
摘要点击次数: 89
全文下载次数: 38
中文摘要:
      随着各种新技术的出现,传统的恶意代码的识别和分类技术存在着检测率瓶颈、实时监测效率不高的问题,为了提高准确率,提出了一种基于图像纹理指纹特征与深度学习神经网络结合的分类方法。该方法首先将数据集中恶意代码的二进制文件建模为灰度图,采用改进的灰度共生矩阵提取出恶意代码中的指纹特征图像,并选择不同步长扩展样本量,然后将该指纹特征图像作为输入数据集并采用卷积神经网络模型中进行分类训练。结果表明,该方法可以有效地分类恶意代码,准确率可达96.2%,并在泛化测试中取得了较好的效果。
英文摘要:
      With the emergence of various new technologies, there are some problems of the low efficiency of real-time detection in traditional malware identification and classification techniques. In order to improve the accuracy, this paper proposes a classification method based on image texture fingerprint features combined with deep learning neural networks. First, the binary file of malware in the data set is modeled as a grayscale image, and improved grayscale co-occurrence matrix is used to extract the fingerprint feature image in the malware. The matrixs of different steps are selected to expand the sample size, then the fingerprint feature image is taken as the input data set and the convolutional neural network model is used for classification training. The results show that this method can effectively classify malware with an accuracy rate of 96.2%, and has achieved good results in generalization testing.
查看全文  查看/发表评论  下载PDF阅读器
关闭
你是第33883522位访问者
版权所有:科学技术与工程编辑部
主管:中国科学技术协会    主办:中国技术经济学会
Tel:(010)62118920 E-mail:stae@vip.163.com
京ICP备05035734号-4
技术支持:本系统由北京勤云科技发展有限公司设计

京公网安备 11010802029091号