Research
Shift-aware modeling
I study why EEG pipelines fail across sessions and how to diagnose, benchmark, and adapt them without hand-wavy heuristics.
Data Scientist · Statistical Modeling · Brain-Computer Interfaces
Building robust learning systems for EEG, time series, and distribution shift.
I am a Ph.D. Candidate at UMass Boston working with Prof. David Degras. My work combines statistical modeling, BCI methodology, and scientific software in R and Python.
R / Rcpp Python Statistical Learning EEG Signal Processing Domain Adaptation Drift Diagnostics Evaluation Rigor
Engineering
Scientific tooling
I build R and Python packages that make reproducible experimentation, feature extraction, and domain adaptation easier to run at scale.
Practice
Decision-ready outputs
I care about methods that are interpretable, benchmarkable, and usable in real analysis workflows rather than only looking good on paper.
Selected Work
Research directions
Drift-Feature-Performance Decomposition in MI-EEG via Structured Geometric Modeling
Working Paper (Under Review)A geometric framework that diagnoses BCI performance degradation by separating signal drift into raw sensor variability and feature-space distortions.
Cross-Session BCI Transfer: A Matched Comparison of Global and Selective Pooling
Working Paper (Under Review)A matched comparison of global vs. selective source-session pooling strategies for cross-session EEG transfer, with confidence-interval gating for robust strategy selection.
Tensor Multiple Canonical Correlation Analysis (TMCCA): A Framework for Multi-View High-Dimensional Data Integration
Working Paper (In Preparation)A novel tensor-based statistical framework extending MCCA to high-dimensional datasets, preserving structural information in multi-view neuroimaging analysis.
Balancing Performance and Stability in Motor Imagery EEG Decoding: A Linear-First Approach
Working Paper (Under Review)Proposing a 'Linear-First' decision rule using Paired Non-Inferiority Tests (TOST) to balance decoding accuracy against computational cost.
Projects
Recent builds
Web App: MSDA-Bench
Application DeveloperAn interactive Streamlit dashboard for benchmarking multi-source domain adaptation strategies in cros...
Python Package: DA4BCI
Python Package DeveloperPython port of the DA4BCI R package — a unified framework for domain adaptation in EEG-based BCI w...
Desktop App: NeuroStream
Application DeveloperA real-time BCI motor imagery streaming application with four selectable pipelines (CSP, FBCSP, TS+LD...
R Package: BCIFeatR
R Package DeveloperA feature-engineering toolkit for EEG-based BCI pipelines with 9 extraction methods, Riemannian geometr...
Writing
Latest notes
Practical Guide: Distribution Distance Metrics for EEG Domain Adaptation
A practical note on Wasserstein, MMD, and Energy Distance for quantifying session-to-session shift in EEG pipelines.
From Matrix ICA to Tensor ICA: Architectures and Decomposition Choices
A concise bridge from classical ICA to multilinear/tensor ICA design choices.
ICA Objective Functions in One Place: Kurtosis, Negentropy, MI, and MLE
A compact map of the main ICA objective functions and their algorithmic implications.
Research Note: Markov-Switching State-Space Models for Neural Decoding
A working note on using regime-switching linear state-space models for non-invasive brain-signal decoding.
Contact
Interested in BCI, statistical modeling, or scientific software work?
I am looking for roles where rigorous modeling and production-quality research tooling both matter. The fastest way to reach me is email.