The VinUni-Illinois Smart Health Center (VISHC) is delighted to celebrate a new academic achievement by Nguyen Dai Nghia, a VinUniversity alumnus and VISHC Fellow, whose recent paper on continual learning has been recognized as a Featured paper at Transactions on Machine Learning Research (TMLR) and attached to ICML 2026 in Seoul, South Korea.
Nguyen Dai Nghia – Author of the paper
Currently a second-year Ph.D. student in Computer Science at the University of Illinois Urbana-Champaign, Nghia focuses his research on machine learning and causal inference, with an emphasis on applications in healthcare. His academic journey has taken him from VinUniversity to one of the leading computer science research environments in the United States, where he continues to explore challenging problems at the intersection of artificial intelligence and health.
As one of the first-cohort students at VinUniversity, Nghia had the opportunity to explore different academic pathways before gradually developing a strong interest in research. He initially worked on virtual reality before shifting toward machine learning, particularly continual learning, and was later awarded the VISHC Fellowship to further pursue his Ph.D. at UIUC.
Nghia’s recent paper on continual learning has been recognized as a Featured paper at Transactions on Machine Learning Research (TMLR) and attached to ICML 2026 in Seoul, South Korea. This recognition is particularly meaningful, as TMLR is a peer-reviewed venue for disseminating machine learning research, where accepted papers may receive special certifications such as Featured and Outstanding as public endorsements of their quality and contribution. The work’s connection to ICML 2026 further highlights its academic significance, as the International Conference on Machine Learning (ICML) is one of the premier international venues for machine learning research, bringing together leading researchers working at the frontier of artificial intelligence and machine learning. ICML 2026 will be held from July 6 to 11, 2026, at the COEX Convention & Exhibition Center in Seoul, South Korea.
The featured paper proposes a new perspective on continual learning, a field that studies how machine learning models can learn new tasks over time without forgetting previously acquired knowledge. This challenge, commonly known as catastrophic forgetting, remains one of the central problems in building adaptive and reliable AI systems.
The work approaches this problem by retrospectively correcting catastrophic forgetting through small adapters that can be sequentially chained during inference. By design, the method separates the learning of new tasks, known as plasticity, from the preservation of old knowledge, known as stability. This enables greater flexibility in managing the plasticity-stability trade-off, one of the key challenges in continual learning research.
Figure illustrating how an AI model can “remember” old knowledge after learning new tasks.
Having the work attached to ICML 2026 also provides Nghia and his collaborators with an opportunity to present their ideas to the international machine learning research community. For emerging researchers, such academic platforms are valuable not only for sharing technical contributions but also for receiving feedback, exchanging perspectives, and engaging with new directions in the field.
Beyond this publication, Nghia’s current research direction at UIUC focuses on machine learning and causal inference, especially in healthcare applications. He is particularly interested in open and challenging problems where advances in causal inference, neural networks, and temporal data such as time series can create meaningful impact for adjacent fields, including smart health and healthcare innovation.
“When I was first introduced to causal inference, I saw many interesting but also challenging questions that, once addressed, could positively impact adjacent fields, such as health care.”
Nghia also emphasized the important role of mentorship and collaboration throughout his research journey. In interdisciplinary fields, it is difficult to conduct impactful research alone. The support of VISHC faculty members, principal investigators, and mentors has helped him explore meaningful research directions, broaden his academic perspective, and engage with complex problems that require expertise from multiple areas.
“It is difficult to do research alone and even more so in an interdisciplinary context.”
Nghia’s progress reflects how early research exposure, mentorship, and international collaboration can help young scholars grow into more independent researchers. From his first years at VinUniversity to his current Ph.D. journey at UIUC, his development shows the potential of VISHC-affiliated scholars to contribute to global scientific communities in artificial intelligence, machine learning, and smart health.
His achievement also points to a broader story of young Vietnamese researchers entering international academic spaces with meaningful and technically rigorous work. Rather than presenting research as a linear path, Nghia’s journey shows that academic growth often comes from exploration, uncertainty, collaboration, and the ability to keep learning through new challenges.
“Figuring it out on the way and accepting what comes next is part of the journey.”
