Marina Decker's Research Page

Position

PhD. student | 2026-

Contact

marina.decker@polymtl.ca

Github

B-431.1 Main Building

Ecole Polytechnique

Montreal, CANADA

Education

Bachelor's and Master's in Applied Mathematics | 2019–2025
UFR of Mathematics, Computer Science and Mechanics, University of Metz, France

Research

Project

Title: Hierarchical Prediction of Mutually Exclusive Bridge Condition States Using Network-Scale Censored Data

Summary:
A hierarchical probabilistic approach is proposed to model and predict the condition states of bridges using network scale inspection data. The method is designed for discrete and mutually exclusive states, where the best visual condition corresponds to the highest value and the poorest condition to the lowest. It incorporates the uncertainty inherent in visual inspection processes and enables bridges to share statistical information through common physical and environmental characteristics such as construction material, age, and geographical context. The approach seeks to describe how structural conditions evolve over time and to lay the groundwork for tools that support maintenance planning and prioritization. The analysis is conducted on data from the bridge network in Quebec, with the longer term goal of adapting and applying the methodology to other regional systems.