A Polynomial Time MCMC Method for Sampling from Continuous DPPs

Abstract

We study the Gibbs sampling algorithm for continuous determinantal point processes. We show that, given a warm start, the Gibbs sampler generates a random sample from a continuous $k$-DPP defined on a $d$-dimensional domain by only taking $\text{poly}(k)$ number of steps. As an application, we design an algorithm to generate random samples from $k$-DPPs defined by a spherical Gaussian kernel on a unit sphere in $d$-dimensions, ${\mathbb{S}}^{d-1}$ in time polynomial in $k,d$.

References in corpus (11)

• Determinantal point processes for machine learning
• Determinantal Processes and Independence
• Eynard-Mehta theorem, Schur process, and their pfaffian analogs
• Determinantal point process models and statistical inference : Extended version
• Statistical properties of determinantal point processes in high-dimensional Euclidean spaces
• Efficient Sampling for k-Determinantal Point Processes
• Approximate Inference in Continuous Determinantal Point Processes
• Fast Mixing for Discrete Point Processes
• Learning Determinantal Point Processes with Moments and Cycles
• Exact Sampling from Determinantal Point Processes
• Fast Mixing Markov Chains for Strongly Rayleigh Measures, DPPs, and Constrained Sampling