Although machine learning has achieved unprecedented success over a variety of tasks and across different domains, because of its black-box nature, concerns arise as to whether it can be deployed safely, especially in security-critical environments. People recently have found the state-of-art machine learning models have lacks trust in a lot of aspects. For example, deep neural networks lacks interpretability and are vulnerable to attacks both in the training time and inference time. In this project, we aim to explore research direction in emerging research areas related to the broader study of security and privacy in machine learning.