Abstract: In an interconnected world where machine learning technologies are becoming the backbone of various systems, the imperative for robust and ethical AI solutions is paramount. In this seminar, Dr. Pablo Rivas will explore this critical issue through the lens of two ongoing research projects. The first part of the talk will discuss the use of natural language processing to understand criminal activities in online consumer-to-consumer marketplaces. Dr. Rivas will outline the algorithms and techniques that have been developed to analyze data, aiming to create safer online platforms. The second part will focus on fortifying machine learning models against adversarial attacks. Specifically, the seminar will delve into efforts to improve model robustness and the effects on performance. By the end of the seminar, attendees will have a better grasp of the role large language models play in online safety and will gain insights into approaches for enhancing the safety of deep learning models.
Bio: Dr. Pablo Rivas is an assistant professor of computer science with a focus on both classic and quantum machine learning, natural language processing, and AI ethics. With over a decade of multi-disciplinary experience, he is a senior member of IEEE and ACM. Dr. Rivas has an extensive publishing record in AI and is at the forefront of ethical AI development, actively collaborating with the IEEE Standards Association. He is spearheading the launch of the NSF IUCRC for Standards and Ethics in AI and serves as an associate editor for the IEEE Transactions on Technology and Society.