Author : Rupali Malviya and Brajesh K. Umrao

The power of Machine learning to rapidly gain through experience and evolve with changing and complex situations has helped it become an essential tool for the security of computers. However, its this pliancy is also a vulnerability. It makes the machine learning systems susceptible to attacks. The attackers can exploit machine learning systems because of its nature of adaptability. In this paper we try to analyze different attacks against machine learning systems and their solutions. We examine the contemporary work in this field and present a survey of potential attacks against machine learning systems and the defenses against these attacks.

Keywords: Machine learning, Security, Attacks, Defenses.

Article published in International Journal of Current Engineering and Technology, Vol.4, No.6 (Dec-2014)