Linear / WLS / Robust Regression • Lasso / Ridge / SCAD • Splines • Cox Regression
I deliver end-to-end workflows in R for classical and modern regression: clean data pipelines, defensible model choice, strong diagnostics, and publication-quality figures. I’m fluent with OLS, WLS (precision weights/variance functions), and robust regression (Huber/Tukey M-estimators, quantile regression) to address heteroskedasticity, leverage, and outliers.
Regularization & selection: LASSO, Ridge, Elastic Net (glmnet), information criteria (AIC/BIC), cross-validation, stability selection; SCAD/MCP via ncvreg; grouped penalties; post-selection refits and inference. I provide coefficient paths, tuning curves, and nested model comparisons.
Nonlinearity & smooths: spline regression (splines/splines2), natural/cubic/bs(), knots and df control, GAM with mgcv, interactions and partial effects, and shape-constrained fits when needed. Diagnostics include residual patterns, influence, VIF/collinearity, specification (RESET), and predictive checks.
Survival analysis: Cox proportional hazards (survival/survminer), stratification, time-varying covariates, proportionality tests (Schoenfeld), baseline hazards, and AFT models; cross-validated C-index and calibration.
Deliverables: clean R scripts, tidy outputs, coefficient/diagnostic plots, and a concise memo for reports or theses. I’m very familiar with these methods and can provide hands-on R programming help for coursework, projects, and research.
Get help: engagements start at USD $150; fixed quotes follow a brief review of your data and scope.
