On two ways to use determinantal point processes for Monte Carlo integration
The work investigates two determinantal point process (DPP) estimators for Monte Carlo integration: the Bardenet-Hardy (BH) estimator and the Ermakov-Zolotukhin (EZ) estimator. These methods use DPPs to enforce repulsion between sampling points, leading to estimators with faster convergence rates than classical Monte Carlo for smooth functions and providing theoretical links to Fourier coefficients.