Home

Awesome

计算机安全工程师的机器学习

这是？

这是本人之前工作中读到的一些论文, 所以假设你已经具有了基础的渗透能力和机器学习能力。机器学习(Or深度学习)在安全领域可以应用的地方还是很多的。例如:

image

这些检测与绕过大部分无非是与文本相关(也许在内存里中，也许在文件中)的处理，所以可算是NLP在安全领域的应用,同时也可以看到,将二进制文件转换为灰度图可以用来检测病毒,所以这个可以算是CV的迁移应用。同样，在绕过检测这个环节里，有可能会用到RL的一些知识。

格式如下:

**[序号]** 作者. "名称", [[pdf]](链接),年份 **(描述,应用场景)** 推荐星级 :star::star::star::star::star:

目录

正文

Webshell Detection

[1] Ye Zhang, Byron Wallace. "A Sensitivity Analysis of (and Practitioners' Guide to) Convolutional Neural Networks for Sentence Classification", [pdf],6 Apr 2016 (文本分类，检测webshell) 推荐星级:star::star::star::star::star:

[2] Yoon Kim. "Convolutional Neural Networks for Sentence Classification" [pdf],3 Sep 2014 (文本分类，检测webshell,XSS等) 推荐星级:star::star::star::star::star:

[3] Johannes Dahse. "Static Detection of Complex Vulnerabilities in Modern PHP Applications" [pdf], 02 Feb 2016 (AST，检测webshell) 推荐星级:star::star::star::star::star:

DGA Detection

[1] Hyrum S. Anderson, Jonathan Woodbridge, Bobby Filar. "DeepDGA: Adversarially-Tuned Domain Generation and Detection" [pdf],6 Oct 2016 (生成对抗网络，DGA检测) 推荐星级:star::star::star::star::star:

[2] Jonathan Woodbridge, Hyrum S. Anderson, Anjum Ahuja, Daniel Grant. "Predicting Domain Generation Algorithms with Long Short-Term Memory Networks" [pdf],2 Nov 2016 (LSTM,DGA检测) 推荐星级:star::star::star::star::star:

Malware Detection

[1] "DeepSign: Deep Learning for Automatic Malware Signature Generation and Classification" [pdf] (病毒签名生成和检测) 推荐星级:star::star::star::star::star:

[2] Hyrum S. Anderson, Anant Kharkar,Bobby Filar. "Evading Machine Learning Malware Detection" [pdf],22-27 July 2017 (绕过病毒检测) 推荐星级:star::star::star::star::star:

[3] L. Nataraj, S. Karthikeyan,G. Jacob. "Malware Images: Visualization and Automatic Classification" [pdf],2011 (二进制可视化，病毒检测) 推荐星级:star::star::star::star::star:

[4] L. Nataraj, Vinod Yegneswaran,Phillip Porras,Jian Zhang. "A Comparative Assessment of Malware Classification using Binary Texture Analysis and Dynamic Analysis" [pdf] (二进制可视化，图像分类，病毒检测) 推荐星级:star::star::star::star::star:

[5] Kolter, Maloof. "Learning to detect malicious executables in the wild" [pdf],2004 (NLP, n-grams of byte codes,静态分析,病毒检测) 推荐星级:star::star::star::star::star:

[6] Microsoft. "Microsoft portable executable and common object file format specification", [pdf],2013 (微软PE文件的格式说明,需要了解的基础知识) 推荐星级 :star::star::star::star::star:

[7] J. Saxe and K. Berlin. "Deep neural network based malware detection using two dimensional binary program features.In Malicious and Unwanted Software (MALWARE)", [pdf],3 Sep 2015 (DNN套路病毒检测，面对未知病毒也可检测) 推荐星级 :star::star::star::star::star:

IDS

[1] 金波,林家骏,王行愚. "入侵检测技术评述[J]. 华东理工大学学报", 21 09 2017 推荐星级 :star::star::star:

Password

[1] Briland Hitaj, Paolo Gasti, Giuseppe Ateniese, Fernando Perez-Cruz "PassGAN A Deep Learning Approach for Password Guesssing", [pdf],1 Sep 2017 (GAN网络 破解密码) 推荐星级 :star::star::star::star::star:

Attack

[1] Alex Graves, "Generating Sequences With Recurrent Neural Networks", [pdf], 5 Jun 2014 (文本生成必备论文，RNN,作为LSTM生成XSS必读论文) 推荐星级 :star::star::star::star::star:

[2] Hila Peleg, Patrice Godefroid,Rishabh Singh, "Learn&Fuzz: Machine Learning for Input Fuzzing", [pdf], 2017 (机器学习和Fuzzing ) 推荐星级 :star::star::star::star::star:

资源

资源再多,也没什么用。以如今互联网的发展程度，找资源可以说是轻而易举。稍微动动脑子,都能拿的到。但是关键在于,有和用是两码事。本来并不想在资源这里列很多东西。但觉得不列出来是不完整的,这些知识都是相辅相成的。但又怕列多了误导别人(说的好像真有人来看似的)。因此,暂且仅挑几本书放在下面吧。