Vafa's research focused on leveraging machine learning for end-user security and privacy protection. This was achieved by strengthening the security and privacy defenses over the network stack: in application layer I worked on applying machine learning to build a human-centered anti-fingerprinting defense. In mid-layers I work on the MUD standard for protecting IoT devices through network-microsegmentation., and close to physical layer, I worked on deep learning applications in binary analysis for facilitating reverse engineering of stripped IoT firmware binaries.