摘要
网络环境下，恶意程序对网络安全的威胁日益严重。恶意程序是一种不经使用者同意就可以将电脑程式植入使用者电脑中，造成使用者资料的完整性、机密性与可用性。其中包括病毒，木马，蠕虫，间谍软件等。黑客可以通过网络攻击，恶意附件，恶意网站等多种途径入侵用户的电脑。无论是企业还是个体用户，都面临着不可忽视的危险，因而，对其进行有效的检测与防护显得尤为重要。
关键词：机器学习，恶意程序探测，特征抽取
Abstract
Along with the development of the Internet, the harmfulness of the malicious program is growing. Malicious programs are malicious programs that are secretly installed on users' computers and threaten the integrity, confidentiality, and accessibility of users' data. There are a lot of different kinds of malware, including viruses, trojans, worms, spyware, etc. There are many ways to hack into a user's computer, for example, by means of cyber attacks, malicious attachments, malicious sites, and so on. The threat of malicious program can not be neglected for business enterprises and individual users. Therefore, the research of effective detection and defence strategy is of great importance.
Key words: Machine Learning, Malicious Program Detection and Feature Extraction
目录
摘要 1
Abstract 1
1前言 3
1.1调查背景 3
1.2调查研究的重要性 3
2 恶意程序的演变与分类 3
2.1 什么是恶意程序 3
2.2恶意程序研究进展 4
3 利用机器学习技术对恶意程序进行探测 4
3.1机器学习（机器学习）的理论基础 4
3.2数据预处理技术 5
4防卫战略的探讨 5
4.1 针对恶意软件的防护措施 5
4.2 以机器学习为基础的安全防护技术 6
5结论 6
参考文献 7
