A Georgia Tech Ph.D. student has been selected as an Apple Scholar for work done to ensure machine learning (ML) systems are easy to use and understand for researchers, developers, and everyday end-users.
Zijie Jay Wang is a recipient of the 2023 Apple Scholars in AI/ML PhD fellowship. The fellowship recognizes the success of Wang’s research efforts to make artificial intelligence (AI) and ML systems more transparent and accessible.
As the power and complexity of these systems continue to advance at breakneck speed, his advisor says Wang’s work is vitally important.
“Jay's research connects humans and ML systems. His work on easy-to-use and interactive interfaces is crucial to making ML more interpretable, accessible, and reliable,” said Polo Chau, School of Computational Science and Engineering associate professor.
To help people better connect with these systems, Wang leverages his expertise, not only in ML, but also in data visualization. He’s developed several interactive visualization tools that are designed to help technical and nontechnical people understand deep learning, neural networks, and other ML-related topics. These include:
CNN Explainer, an open-source tool developed for deep-learning beginners. Since its release in July 2020, more than 180,000 global visitors have used the tool.
GAM Changer, which empowers users in healthcare, finance, or other domains to edit ML models to include knowledge and values specific to their domain, which improves reliability. This tool won a best paper award at the 2021 Conference on Neural Information Processing Systems. It has also been integrated into Microsoft's Interoperability Library.
DiffusionDB, a first-of-its-kind large-scale dataset that lays a foundation to help people better understand generative AI. This work could lead new research in detecting deepfakes and designing human-AI interaction tools to help people more easily use these models.
Building upon his success, Wang has two accepted papers at the upcoming ACM CHI Conference on Human Factors in Computing Systems.
One of these, GAM Coach: Towards Interactive and User-centered Algorithmic Recourse, describes an interactive ML tool that could be used to help people who have been rejected for a loan by automatically letting an applicant know what’s needed for them to receive loan approval.
The Apple Scholars in AI/ML PhD fellowship program will enable Wang to explore more work like this, as well as more high-risk/high-reward research opportunities.
“It is a tremendous privilege to be awarded this fellowship, and I am excited about the opportunity to collaborate with researchers at Apple in my research projects. By partnering with Apple researchers, I believe it will greatly amplify real-world applicability and impacts of my work,” said Wang.
This is not the first time Wang has been recognized with a high-profile fellowship. He was recognized last year as a recipient of the 2022 J.P. Morgan AI Ph.D. Fellowship