For the class final project:
- Reviewed the paper "Accelerated Gibbs sampling of normal distributions using matrix splittings and polynomials" by C. Fox and A. Parker (2017) and implemented the proposed generalized and accelerated Gibbs samplers algorithm using the Julia programming language.
- The algorithm we implemented was found to be more memory-efficient, using over ten times less memory compared to iterative solvers and was also faster than the Julia iteration solver package.
- We also expanded the scope of the project by exploring the authors' related series of research, implementing PCG and Cheby accelerated samplers, and additional related forms.