Approximate Inference in Continuous Determinantal Point Processes
Raja Hafiz Affandi, Emily B. Fox, Ben Taskar

Citation
Raja Hafiz Affandi, Emily B. Fox, Ben Taskar. "Approximate Inference in Continuous Determinantal Point Processes". NIPS 2013, Neural Information Processing Systems Foundation, 5, December, 2013; Presented at the NIPS 2013 Conference. .

Abstract
Determinantal point processes (DPPs) are random point processes well-suited for modeling repulsion. In machine learning, the focus of DPP-based models has been on diverse subset selection from a discrete and finite base set. This discrete setting admits an efficient sampling algorithm based on the eigendecomposition of the defining kernel matrix. Recently, there has been growing interest in using DPPs defined on continuous spaces. While the discrete-DPP sampler extends formally to the continuous case, computationally, the steps required are not tractable in general. In this paper, the authors present two efficient DPP sampling schemes that apply to a wide range of kernel functions: one based on low rank approximations via Nystrom and random Fourier feature techniques and another based on Gibbs sampling. They demonstrate the utility of continuous DPPs in repulsive mixture modeling and synthesizing human poses spanning activity spaces.

Electronic downloads

(No downloads are available for this publication.)

Citation formats

• HTML

  Raja Hafiz Affandi, Emily B. Fox, Ben Taskar. <a
  href="http://www.terraswarm.org/pubs/214.html"
  >Approximate Inference in Continuous Determinantal Point
  Processes</a>, NIPS 2013, Neural Information
  Processing Systems Foundation, 5, December, 2013; Presented
  at the <a
  href="http://nips.cc/Conferences/2013/"
  >NIPS
  2013 Conference</a>. .

• Plain text

  Raja Hafiz Affandi, Emily B. Fox, Ben Taskar.
  "Approximate Inference in Continuous Determinantal
  Point Processes". NIPS 2013, Neural Information
  Processing Systems Foundation, 5, December, 2013; Presented
  at the <a
  href="http://nips.cc/Conferences/2013/"
  >NIPS
  2013 Conference</a>. .

• BibTeX

  @inproceedings{AffandiFoxTaskar13_ApproximateInferenceInContinuousDeterminantalPointProcesses,
      author = {Raja Hafiz Affandi and Emily B. Fox and Ben Taskar},
      title = {Approximate Inference in Continuous Determinantal
                Point Processes},
      booktitle = {NIPS 2013},
      organization = {Neural Information Processing Systems Foundation},
      day = {5},
      month = {December},
      year = {2013},
      note = {Presented at the <a
                href="http://nips.cc/Conferences/2013/"
  >NIPS
                2013 Conference</a>. },
      abstract = {Determinantal point processes (DPPs) are random
                point processes well-suited for modeling
                repulsion. In machine learning, the focus of
                DPP-based models has been on diverse subset
                selection from a discrete and finite base set.
                This discrete setting admits an efficient sampling
                algorithm based on the eigendecomposition of the
                defining kernel matrix. Recently, there has been
                growing interest in using DPPs defined on
                continuous spaces. While the discrete-DPP sampler
                extends formally to the continuous case,
                computationally, the steps required are not
                tractable in general. In this paper, the authors
                present two efficient DPP sampling schemes that
                apply to a wide range of kernel functions: one
                based on low rank approximations via Nystrom and
                random Fourier feature techniques and another
                based on Gibbs sampling. They demonstrate the
                utility of continuous DPPs in repulsive mixture
                modeling and synthesizing human poses spanning
                activity spaces.},
      URL = {http://terraswarm.org/pubs/214.html}
  }
  

Posted by Barb Hoversten on 15 Nov 2013.