Abstract: Clinical medical devices and healthcare IT systems have been profoundly affected by an evolving cybersecurity threat landscape. And as federal regulators and the healthcare industry have worked together to address these threats, a new category of cybersecurity science has emerged, one that combines cybersecurity disciplines, medical science, and regulatory science to create a new breed of cybersecurity professional. This presentation will examine the unique status of medical systems in the realm of cybersecurity, their common attack surfaces, and the challenges involved in safeguarding these systems. The future of the medical cybersecurity market will be discussed, along with the rapidly expanding career opportunities for cybersecurity professionals in this field.

Title: Analyzing and Securing Software via Robust and Generalizable Learning

Abstract: Software is powering every aspect of our society, but it remains plagued with errors and prone to critical failures and security breaches. Program analysis has been a predominant technique for building trustworthy software. However, traditional approaches rely on hand-curated rules tailored for specific analysis tasks and thus require significant manual effort to tune for different applications. While recent machine learning-based approaches have shown some early promise, they, too, tend to learn spurious features and overfit to specific tasks without understanding the underlying program semantics. In this talk, I will describe my research on building machine learning (ML) models toward learning program semantics so they can remain robust against transformations in program syntax and generalize to various program analysis tasks and security applications. The corresponding research tools, such as XDA, Trex, StateFormer, and NeuDep, have outperformed commercial tools and prior arts by up to 117x in speed and by 35%in precision and have helped identify security vulnerabilities in real-world firmware that run on billions of devices. To ensure the developed ML models are robust and generalizable, I will briefly describe my research on building testing and verification frameworks for checking the safety properties of deep learning systems.

Kexin Pei is an incoming Neubauer Family Assistant Professor at the Department of Computer Science, University of Chicago. He received his PhD at Department of Computer Science, Columbia University. His research lies at the intersection of security, software engineering, and machine learning, with a focus on building machine-learning tools that utilize program structure and behavior to analyze and secure software. His research has received the Best Paper Award in SOSP, an FSE Distinguished Artifact Award, been featured in CACM Research Highlight, and won CSAW Applied Research Competition Runner-Up. He was part of the LearningForCode team at Google DeepMind, building program analysis tools based on large language models.

Seminar topic TBA.

Amanda Hilliard is a Senior Lecturer for the JHU Center for Leadership Education. Before joining the Center for Leadership Education, Hilliard had previously taught writing and communication courses abroad in South Korea, Vietnam, and Ecuador, and in the U.S. in Georgia, Texas, and Arizona. She has worked with the U.S. State Department as an English Language Fellow and a webinar presenter, and as an ESL instructor with the U.S. Department of Defense. Having served on the board of AZTESOL and organized numerous conferences, Hilliard strongly believes in professional development for educators and regularly conducts teacher training. She has published and presented on a wide variety of topics, including grammar instruction, data-driven language learning, metaphor, culture, and language policy. Hilliard aims to help students communicate effectively for diverse audiences in today’s increasingly multicultural, globalized, and digital world.