Trevor L. Wrobleski
| Home | Research | Projects & Awards |
Welcome
I am a Statistics PhD candidate at the London School of Economics, interested in using Bayesian methods to solve uncertainty quantification problems in risk assessment, toxicology/pharmacology, and energy markets. My advisors are Dr. Sara Geneletti and Dr. Francesca Panero. My research focuses on developing robust statistical models to improve decision-making in toxicology, pharmacology/biotech, energy markets, and risk systems under uncertainty.
Research Interests
Statistical Methodology & Theory
- Bayesian Model Averaging & Stacking: Predictive distribution combination in M-open settings, scoring rules (log-score, CRPS), and coherency vs. predictive accuracy trade-offs.
- Optimal Experimental Design: Locally D-optimal and c-optimal designs for non-linear models, resource allocation strategies (N vs. G), and design robustness under overdispersion.
- Robust Inference: Handling model misspecification, variance propagation in poly-k survival adjustments, and quantifying bias in finite-sample BMD estimation.
Quantitative Risk Assessment
- Dose-Response Modeling: Hierarchical modeling of heterogeneity (chemicals/studies), benchmark dose (BMD) estimation, and derivation of empirical priors for regulatory science.
- Toxicology & Carcinogenicity: Analyzing large-scale bioassay databases (CPDB), accounting for litter effects and differential mortality, and improving points of departure (PoD) for FDA/EMA submissions.
- Energy Markets & Commodity Risk: Probabilistic forecasting under volatile and illiquid market conditions, portfolio risk quantification, and Bayesian approaches to uncertainty in energy commodity pricing.
Application Areas
Academic Research: Advancing Bayesian model averaging theory, developing robust methods for small-sample inference, creating open-source tools for reproducible risk assessment
Regulatory Science & Strategy: Benchmark dose estimation for FDA/EMA submissions, prior sensitivity analysis for regulatory compliance, optimal study design to maximize approval probability while minimizing cost
Pharmacology & Drug Development: Quantitative due diligence for biotech assets, probability of technical success (PoTS) modeling, adaptive trial design, go/no-go decision frameworks, risk-adjusted NPV analysis for drug development pipelines
Energy Markets & Risk: Portfolio optimization under uncertainty, probabilistic forecasting for energy commodity markets, quantifying risk in volatile and illiquid markets
Healthcare Analytics: Causal inference from real-world evidence, predictive modeling for patient recovery trajectories, synthetic control methods for post-market surveillance
Education
- PhD Candidate, Statistics – London School of Economics
- MSc, Statistics (Research) – London School of Economics
- MMSc, Global Affairs, Economics and Business – Tsinghua University, Schwarzman College (Bloomberg Fellow)
- BA, Public Health – Johns Hopkins University
Skills
| Category | Proficiencies |
|---|---|
| Statistical Methods | Bayesian inference, Optimal experimental design, Hierarchical (mixed-effects) modeling, Survival analysis, MCMC |
| Programming | R (Tidyverse, Shiny), Python (NumPy, Pandas, Flask), SQL, Git |
| Bayesian & ML | Stan, PyMC, TensorFlow, PyTorch, Scikit-learn |
| Languages | English (Native), French (Intermediate), Chinese (Mandarin/普通话) |
About Me
Outside of research, I enjoy running, tennis, hiking, reading, writing, the ballet and opera, and playing piano.
You can contact me from my university page.