MCMC methods for rare events and non-convex supports
Developing novel sampling methods, with applications to rare event sampling and stochastic optimization.
See below for the list of related publications and preprints.
Related
Publications
Uncovering Load-Altering Attacks Against $N-1$ Secure Power Grids: A Rare-Event Sampling Approach (2023)
M. Goodridge, S. Lakshminarayana, A. Zocca
Submitted to IEEE Transactions on Power Systems
How to identify the most impactful dynamic load-altering attacks?
Analysis of Cascading Failures Due to Dynamic Load-Altering Attacks (2023)
M. Goodridge, A. Zocca, S. Lakshminarayana
IEEE SmartGridComm 2023
What are the features of the most impactful dynamic load-altering attacks?
Hopping between distant basins (2022)
M. Goodridge, J. Moriarty, J. Vogrinc, A. Zocca
Journal of Global Optimization
How to improve the Basin Hopping method for global optimization with nonlocal random perturbations?
A Metropolis-class sampler for targets with non-convex support (2021)
J. Moriarty, J. Vogrinc, A. Zocca
Statistics and Computing
How to efficiently sample from target densities with non-convex support? Introducing the Skipping Sampler
Frequency violations from random disturbances: an MCMC approach (2018)
J. Moriarty, J. Vogrinc, A. Zocca
2018 Conference on Decision and Control (CDC)
How to efficiently sample rare events in power systems? Introducing the Ghost Sampler