ACSSpack - ACSS, Corresponding ACSS, and GLP Algorithm
Allow user to run the Adaptive Correlated Spike and Slab
(ACSS) algorithm, corresponding INdependent Spike and Slab
(INSS) algorithm, and Giannone, Lenza and Primiceri (GLP)
algorithm with adaptive burn-in. All of the three algorithms
are used to fit high dimensional data set with either sparse
structure, or dense structure with smaller contributions from
all predictors. The state-of-the-art GLP algorithm is in
Giannone, D., Lenza, M., & Primiceri, G. E. (2021,
ISBN:978-92-899-4542-4) "Economic predictions with big data:
The illusion of sparsity". The two new algorithms, ACSS
algorithm and INSS algorithm, and the discussion on their
performance can be seen in Yang, Z., Khare, K., & Michailidis,
G. (2024, preprint) "Bayesian methodology for adaptive sparsity
and shrinkage in regression".