Reproduces the work of Ma et al. (2018) characterizing adversarial data subspaces using Local Intrinsic Dimensionality (LID). Leverages recent Topological Data Analysis (TDA) theories and tools to improve the safety of machine learning models, proposes hypotheses on results, and designs numerical verification experiments.
Read More →Research Projects
Research experience and course projects in scientific computing, topological data analysis, coding theory, and mathematical modeling.
Designed a finite difference scheme to solve the Adiabatic Heating Equation (an elliptic equation describing latent release) of the Quasi-Geostrophic Equation Set in a parallelized manner. Leveraged the spectral spherical harmonics method to improve convergence, applying techniques from Scientific Computation (lectured by Alexander Kurganov). Supervised by Dr. Li Dong.
Read More →Mathematical Contest in Modeling 2025
Developed a network model to analyze, simulate, and optimize traffic flows of various transportation shares in Baltimore, MD. Won the Honorable Mention Prize in the MCM-ICM 2025 competition.
Hamming Codes and Golay Codes
Individual study on Hamming Codes and Golay Codes during the Applied Abstract Algebra course taught by Fields Medalist Efim Zelmanov. Presented findings in class.
Proposed a Bayesian pricing and inventory strategy model, optimized via polar coordinate transformations and integrated with machine learning algorithms. Awarded Provincial Third Prize in the China Undergraduate Mathematical Contest in Modeling (CUMCM) 2023.
Read More →