|
|
|
|
Zero-inflated GLMM with hierarchical random effects
Methods: Zero-Inflated Gamma GLMM, random intercepts/slopes, DHARMa diagnostics, Markov chain analysis Data: 142,000+ daily observations across 49 locations Key Features: Rigorous model selection (M0-M6), spatial heterogeneity estimation, temporal dependence quantification |
Scalable Bayesian inference for policy evaluation via INLA
Methods: Bayesian Structural Time Series, Laplace approximation via INLA, causal identification assumptions, interrupted time series design Focus: Efficient Bayesian computation for large time series data, posterior uncertainty quantification |
SPDE approximations for efficient spatial inference
Methods: SPDE approximation, CAR models, spatial random effects, Bayesian hierarchical inference Focus: Geographic autocorrelation, efficient computation for spatial GLMMs, posterior spatial field estimation |
|
End-to-end Bayesian ML pipeline with statistical validation
|
Graph-based retrieval augmented generation for medical research
|
Vision-language models for clinical documentation
|
|
R Statistical Analysis |
Tidyverse Data Wrangling |
ggplot2 Data Visualization |
tidymodels ML Framework |
glmmTMB Generalized Linear Mixed Models |
|
INLA Bayesian Inference |
DHARMa Residual Diagnostics |
Git Version Control |
Quarto Reproducible Docs |
PostgreSQL Database |
|
Python Primary Language |
NumPy Scientific Computing |
Pandas Data Manipulation |
PyTorch Deep Learning |
TensorFlow Neural Networks |
|
Scikit-learn Machine Learning |
Docker MLOps |
Git Version Control |
Jupyter Notebooks |
MongoDB NoSQL |
No activity tracked
|
|
|
|
|