Secure Machine Learning, a Brief Overview

Abstract

The purpose of this article is to give a brief overview on the current work towards the emerging research problem of secure machine learning. Machine learning technique has been applied widely in various applications especially in spam detection and network intrusion detection. Most existing learning schemes assume that the environment they settle in is benign. However this is not always true in the real adversarial decision-making situations where the future data sets and the training data set are no longer from the same population, due to the transformations employed by the adversaries. As more and more machine learning systems are put into use, it is imperative to consider the security of the machine learning system. As a emerging problem, it is attracting more and more researchers' attention. In this article, we present a brief overview on secure machine learning and current progress on developing secure machine learning algorithms.