Code
Code for different projects, including research papers, educational materials and various software tools. Primarily contains Bayesian inference, machine learning and computational mechanics applications.
Research Projects
Decision-theoretic model assessment for structural deterioration monitoring
Code and data to reproduce results from: "On decision-theoretic model assessment for structural deterioration monitoring, (MSSP, 2024)". Implements Bayesian inference using Numpyro (JAX) and implements decision-theoretic utility function calculations for model assessment in structural health monitoring applications.
Hierarchical Bayesian modelling for crack growth monitoring
Implements hierarchical Bayesian modelling using Numpyro (JAX) to estimate crack growth model parameters using crack growth data. Hierarchical models used in the context of multi-task learning to improve inference performance by pooling information from similar structural components. Results compared to typical single-task learning (no pooling).