Shuzhen ZhangResearch Interests
- Unsupervised learning and manifold learning
- Representation learning and embeddings
- Evaluation, robustness, and statistical modeling
- Model interpretability for high-dimensional systems
Research
Themes, active questions, and methods currently guiding my work.
Current Questions I’m Exploring
- How can we quantify embedding reliability across random seeds and data shifts?
- Which diagnostics best predict downstream task performance from low-dimensional structure?
- How should uncertainty estimation be integrated into practical model comparison workflows?
Methods Toolbox
UMAP, t-SNE, Isomap, PCA, clustering diagnostics, GLM/LMM/GEE, bootstrap inference, ablations, and calibration analysis.
Publications & Posters
Publication
- Interactive Visualization of Metric Distortion in Nonlinear Data Embeddings using the distortions Package (bioRxiv preprint, 2025). View on bioRxiv