论文部分内容阅读
回顾过去的几年,机器学习在安全领域有不少应用,但其处境却一直比较尴尬:一方面,机器学习技术在业内已有不少成功的应用,大量简单的重复性劳动工作可以很好地由机器学习算法解决;但另一方面,面对一些“技术性”较高的工作,机器学习技术却又远远达不到标准。和其他行业不同,安全行业是一个比较敏感的行业。比如做一个推荐系统,效果不好的最多也就是给用户推荐了一些他不感兴趣的内容,并不会造成太大损失;而在安全行业就不
Recalling the past few years, machine learning has many applications in the field of security, but its situation has always been more embarrassing: on the one hand, machine learning technology has many successful applications in the industry, a large number of simple and repetitive work can be very good Machine learning algorithms; on the other hand, in the face of some “technical” higher jobs, machine learning techniques are far from standards. Unlike other industries, the security industry is a sensitive industry. For example, to make a recommendation system, the effect is not good is to recommend to the user some of the content that he is not interested in, and will not cause too much damage; in the security industry is not