Vincenzo Coia, PhD

Vincenzo Coia, PhD

Statistician, combining research and practice for probabilistic and risk modelling in the earth sciences.

View CV

About

I am a senior data scientist at BGC Engineering, specializing in practical and innovative probabilistic modeling. I combine advanced statistical methods with real-world applications, focusing on earth systems modelling, extreme events, and communication.

Research Initiatives

My research focuses on expanding how we answer the question, “What’s possible?”–not just “What’s most likely?” I draw on the strengths of both classical statistics and machine learning, building probabilistic models of complex, interdependent systems. My work often centers on understanding extremes in climate and hydrology, where I bridge theoretical innovation with practical application.

The following are research initiatives I’m leading.

Approach

Innovative Statistical Solutions

Leverage a robust foundation in probability theory, extreme value modeling, and machine learning to craft novel methods tailored to real-world challenges.

Comprehensive Uncertainty Analysis

Isolate and communicate multiple sources of uncertainty using a tailored mix of approaches.

Practical Data Science

Ensure that all models are grounded in reality, considering data availability, domain expertise, and end-user needs, avoiding over-engineered solutions.

Reproducible and Efficient Code

Uphold best practices in computation, using version control, defensive programming, and clear documentation to deliver projects that are both reproducible and scalable.

Dynamic Communication and Collaboration

Foster knowledge exchange with domain experts to understand the problem space deeply, ensuring that model results are interpretable, actionable, and presented clearly.

Cutting-Edge Software Development

Design intuitive R packages and Shiny applications, making complex analyses accessible and empowering teams to conduct probabilistic modeling seamlessly.

Featured Work
Scholarly Publications

Articles in scientific journals, proceedings, and preprints.

(2024). Copula-based conditional tail indices. Journal of Multivariate Analysis.
(2024). Probability Distributions of Tailings Dam Breach Volumes by Failure Mode as Part of a Risk Screening-Level Tool. Proceedings of Tailings and Mine Waste 2024.
(2023). Non-Crossing Dual Neural Network: Joint Value at Risk and Conditional Tail Expectation Regression with Non-Crossing Conditions. Available at SSRN 4351877.
(2022). Non-Crossing Dual Neural Network: Joint Value at Risk and Conditional Tail Expectation estimations with non-crossing conditions. IREA–Working Papers, 2022, IR22/15.
(2021). Tail Behavior for Bivariate Distributions Based on Pareto Mixtures. Advances in Statistics - Theory and Applications: Honoring the Contributions of Barry C. Arnold in Statistical Science.
Have a unique challenge or project?

Have a unique challenge or project?

I’m curious about projects that push boundaries and could benefit from a fresh, statistically-driven approach. If that sounds like what you have in mind, I encourage you to connect with me.

Email