Tianyi Gao

Tianyi Gao

Boston University’ 26 | Researcher in Biostatistics & Computational Methods

Email: titigao [at] bu [dot] edu
GitHub: TianyiGao616
LinkedIn: tianyi-gao

About

I’m a recent graduate of Boston University, where I received a B.A. in Mathematics and Computer Science. My research interests broadly lie in applying statistics and computational methods to address health problems, especially those related to complex diseases. I will begin graduate studies in the Master of Science (ScM) in Biostatistics program at the Johns Hopkins School of Public Health.

Research Interests

I’m broadly interested in:

  1. Deep learning, artificial intelligence, and machine learning
  2. Causal inference
  3. Genetics and genomics
  4. Longitudinal methods and survival analysis
  5. Medical image analysis

Research Experience

Below are selected research projects I have worked on:

Boston University, Department of Biostatistics

  • Researcher (2026–Present)
    • Conducting research on Alzheimer’s disease using longitudinal data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI).
    • Studying longitudinal datasets with substantial missingness and evaluating how different imputation methods affect classification model performance.

Boston University, Department of Biostatistics

  • Research Assistant, with Ningyuan Wang and Prof. Ching-Ti Liu (2025–Present)
    • Working on high-dimensional heterogeneous mediation analysis methods
    • Implemented large-scale simulation studies using HPC on BU SCC and computed error metrics to evaluate the performance of different mediation methods
    • Contributing to the development of an R package for the mediation method
    • Currently transferring large-scale datasets to BU SCC for future analyses

Boston University, Department of Mathematics and Statistics

  • Independent Researcher for Undergraduate Honors Thesis, with Dr. Matthew Moore (2025)
    • Modeled ecological and epidemiological systems using ODE-based predator–prey models
    • Extended the classical two-species Lotka–Volterra model to a general n-species framework
    • Incorporated logistic growth terms and analyzed the resulting dynamical system
    • Performed stability and phase-plane analysis of the generalized models
    • Estimated model coefficients using three regression methods and evaluated estimation accuracy using statistical metrics

Boston University, Department of Mathematics and Statistics

  • Research Assistant, with Prof. Mark Kon (2025)
    • Built an ensemble machine learning pipeline for cancer classification using high-dimensional gene expression data (500 samples, 30k genes).
    • Applied SVM as base learners and Random Forest, Decision Tree, and Logistic Regression models as metalearners to identify biologically meaningful gene pathways.
    • Achieved 95% accuracy and extracted top 10–20 genes per meta-learner for biological interpretation.
    • Project code available on GitHub: Repository Link.

Teaching and Mentoring

I am passionate about teaching and enjoy helping students learn and grow. I worked as a Teaching Assistant for Analysis of Variance (MA 416) in Fall 2025. Starting in Fall 2022, I tutored middle and high school students in China in AP Calculus and the TOEFL exam. I also worked as a Peer Mentor at CAS, where I supported first-year students by offering academic guidance and advice on course planning and university life.

Presentations and Conferences

  • Presented my honors thesis, “Modeling Ecological and Epidemiological Systems Using Predator–Prey Dynamics”, at the Undergraduate Research Opportunities Program (UROP) Symposium, Fall 2025.
  • Successfully defended my undergraduate honors thesis before the Department of Mathematics faculty committee in December 2025.

Graph View