This course introduces the foundational machine-learning methods used in modern physics research, combining mathematical concepts with practical computational skills. Students learn how linear algebra, optimization, and statistical modeling support core techniques for analyzing experimental data, interpreting simulations, and building predictive models. The course emphasizes physical intuition, numerical stability, and clear interpretation of results. An overview of PyTorch helps students implement models and apply them to real physics-motivated datasets. By the end of the course, students will understand how key machine-learning tools operate, when they are appropriate for physics problems, and how to evaluate their performance and limitations. Topics covered include linear systems of equations, least-squares methods, linear regression, linear classification, tree-based methods for decision-making, matrix factorization methods (including eigendecomposition, singular value decomposition, and QR), kernel methods, and principal component analysis.
Unlock more information by signing in with your HKUST account.