Research
I work at the intersection of computational neuroscience and neural engineering, with emphasis on methods that can move from exploratory analysis to reliable, real-time use.
Current Project
OSPLAT: Latent Oscillatory Structure Across Time
OSPLAT is my dissertation project in the Pearson Lab at Duke. The goal is to uncover latent oscillatory sources in high-dimensional neural recordings and make those representations useful for downstream inference and interpretation.
Key objectives:
- identify stable latent structure in noisy neural population signals
- support real-time and near-real-time inference workflows
- keep models interpretable enough for scientific and translational use
Related links:
Current Research Focus
Latent Oscillatory Dynamics
I study how to represent oscillatory and time-varying structure in population activity using probabilistic and sparse modeling approaches. The focus is on models that preserve biological meaning while remaining computationally practical.
Real-time Inference Pipelines
I build end-to-end workflows for ingestion, preprocessing, inference, and visualization that can run reliably under practical constraints. This includes reproducible scripting, version control, and deployment-aware engineering decisions.
Cross-Dataset Generalization
I evaluate methods across multiple public neural datasets to test robustness under changing conditions, recording setups, and task structures. The objective is method reliability rather than single-dataset overfitting.
Expanded Interests
- computational neuroscience and neural data decomposition
- neuroengineering tool development
- brain-computer interfaces and translational neurotechnology
- reproducible scientific software and applied ML systems
Research Updates
As new manuscripts, preprints, and presentations are finalized, I will post updates in: