The CIRTech Institute of Technology at Ho Chi Minh City University of Technology (HUTECH) successfully hosted the themed workshop “Scientific Machine Learning” on July 14, aiming to strengthen research collaboration with leading national and international scientists in applied mathematics, scientific computing, and applied machine learning.
Opening the event, Prof. Dr. Le Van Canh, Vice President of HUTECH, shared information about the university’s international research projects and its wide-ranging activities in science and technology. He also expressed his expectations for the workshop: “I hope researchers and students will confidently share and exchange academic insights, contributing to valuable connections that will help further in-depth research development at today’s event.”
Prof. Dr. Le Van Canh encourages researchers and students to engage actively in the workshop
Also speaking at the opening, Prof. Dr. Nguyen Xuan Hung, Director of the CIRTech Institute, provided an overview of the Institute’s activities, highlighting its main directions: advancing fundamental research through prestigious international publications; developing and transferring applied technologies; and training high-quality human resources in science and technology. Emphasizing the potential of the younger generation in driving scientific advancement, he remarked, “The greatest value of education lies in the younger generation’s ability to contribute to the development of the nation.”
Prof. Dr. Nguyen Xuan Hung highlights the potential of the younger generation in scientific research
Kicking off the technical program, Prof. Dr. Bui Thanh Tan, Co-Director of the Center for Scientific Machine Learning at the Oden Institute, Leader of the Pho-Ices Group, and faculty member of the Department of Aerospace Engineering and Engineering Mechanics, as well as the Oden Institute for Computational Engineering and Sciences at The University of Texas at Austin (USA), delivered a presentation titled “Learn2Solve: A Deep Learning Framework for Real-Time Solutions of Forward, Inverse, and UQ Problems.”
Prof. Dr. Bui Thanh Tan presents on “Learn2Solve: A Deep Learning Framework for Real-Time Solutions of Forward, Inverse, and UQ Problems”
The presentation introduced an optimized deep learning framework designed to solve computational problems in real time. The research was driven by the need for deep learning methods capable of efficiently solving complex differential equations, such as the time-dependent scalar probability density function (PDF) equation. In analyzing the method, Prof. Dr. Bui Thanh Tan emphasized key components, including data randomization to improve generalization, the design of loss functions and randomized loss functions to optimize training on uncertain data, and mechanisms for estimating testing errors and defining prediction error at each prediction step. These analyses were supported by theoretical proofs validating the accuracy and numerical stability of the method. The framework’s effectiveness was illustrated through several experimental examples on partial differential equations (PDEs). He concluded, “A valuable computational science study must be grounded in both practical relevance and a strong connection between fundamental research and real-world application.”
The effectiveness of the deep learning framework was illustrated through real-world examples
Following that, Dr. Le Duc Thang, Researcher at CIRTech, presented a report titled “Physics-Informed Neural Networks for Solid Mechanics and Applications.” The presentation highlighted the use of Physics-Informed Neural Networks (PINNs) for solving problems in solid mechanics. PINNs were shown to be especially effective in inverse problems due to their ability to incorporate physical constraints, reduce reliance on experimental data, and enhance prediction accuracy.
Dr. Le Duc Thang presents on “Physics-Informed Neural Networks for Solid Mechanics and Applications”
In the biomedical domain, Dr. Pham The Anh Phu, Lecturer at HUTECH’s Faculty of Information Technology, delivered a presentation titled “Bridging AI, Topological Data Analysis, and Fuzzy Neural Networks for Medical Disease Diagnosis.” The report introduced an interdisciplinary research approach combining AI, Topological Data Analysis (TDA), and Fuzzy Neural Networks (FNNs) to support more accurate diagnosis from complex medical data. The research has been applied in two notable projects: FSpineNet, supporting spinal disease diagnosis at Nguyen Tri Phuong Hospital, and TopoAttn, predicting cardiovascular disease at Thu Duc General Hospital.
Dr. Pham The Anh Phu presents medical research applied in two key projects
The reports presented at the workshop not only demonstrated strong academic value but also highlighted the intersection between computational mathematics, data science, and real-world applications, opening new directions for research in scientific computing and applied machine learning.
The workshop offered an opportunity for faculty, researchers, and students to explore new research directions
The workshop served as an in-depth academic forum, providing faculty, researchers, and students with the opportunity to stay updated on modern research trends and to orient their work toward real-world needs. In doing so, it contributes to enhancing the quality of training and research at the CIRTech Institute in particular and at HUTECH as a whole.
News: Hồng Loan
Photo: Quốc Đạt
Communication Center