A Sparse Bayesian Learning Algorithm for Estimation of Interaction Kernels in Motsch-Tadmor Model
Researchers have developed a new sparse Bayesian learning algorithm to identify interaction kernels within the Motsch-Tadmor model. This method uses a variational framework to reformulate kernel identification as a subspace identification problem. The algorithm incorporates informative priors for regularization and uncertainty quantification, demonstrating accuracy and robustness in numerical experiments. AI