Abstract
We study the computational complexity of zigzag sampling algorithm forstrongly log-concave distributions. The zigzag process has the advantage of notrequiring time discretization for implementation, and that each proposedbouncing event requires only one evaluation of partial derivative of thepotential, while its convergence rate is dimension independent. Using theseproperties, we prove that the zigzag sampling algorithm achieves $\varepsilon$error in chi-square divergence with a computational cost equivalent to$O\bigl(\kappa^2 d^\frac{1}{2}(\log\frac{1}{\varepsilon})^{\frac{3}{2}}\bigr)$gradient evaluations in the regime $\kappa \ll \frac{d}{\log d}$ under a warmstart assumption, where $\kappa$ is the condition number and $d$ is thedimension.