Sketching for large-scale learning of mixture models

@article{Keriven2016SketchingFL,
  title={Sketching for large-scale learning of mixture models},
  author={N. Keriven and Anthony Bourrier and R. Gribonval and P. P{\'e}rez},
  journal={2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  year={2016},
  pages={6190-6194}
}
  • N. Keriven, Anthony Bourrier, +1 author P. Pérez
  • Published 2016
  • Computer Science, Mathematics
  • 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
  • Learning parameters from voluminous data can be prohibitive in terms of memory and computational requirements. We propose a "compressive learning" framework where we first sketch the data by computing random generalized moments of the underlying probability distribution, then estimate mixture model parameters from the sketch using an iterative algorithm analogous to greedy sparse signal recovery. We exemplify our framework with the sketched estimation of Gaussian Mixture Models (GMMs). We… CONTINUE READING
    48 Citations
    Compressive Learning for Semi-Parametric Models
    • 1
    • PDF
    Compressive Statistical Learning with Random Feature Moments
    • 40
    • PDF
    Statistical Learning Guarantees for Compressive Clustering and Compressive Mixture Modeling
    • 1
    • PDF
    Quantized Compressive K-Means
    • 10
    • PDF
    Compressive K-means
    • 31
    • PDF
    Gaussian mixture model decomposition of multivariate signals
    • PDF
    Differentially Private Compressive K-means
    • 6
    • PDF
    Sketched clustering via hybrid approximate message passing
    • 7
    • PDF
    Compressive Learning of Generative Networks
    • PDF
    Inference in Multi-Layer Networks with Matrix-Valued Unknowns
    • 1
    • PDF

    References

    SHOWING 1-10 OF 263 REFERENCES
    Compressive Gaussian Mixture estimation
    • 16
    • PDF
    Scalable Training of Mixture Models via Coresets
    • 105
    • PDF
    Compressive K-means
    • 31
    • PDF
    Toward Learning Gaussian Mixtures with Arbitrary Separation
    • 27
    On learning mixtures of heavy-tailed distributions
    • 55
    • PDF
    On Spectral Learning of Mixtures of Distributions
    • 225
    • PDF