Corpus ID: 3570578

Gaussian quadrature for matrix inverse forms with applications

@inproceedings{Li2016GaussianQF,
  title={Gaussian quadrature for matrix inverse forms with applications},
  author={C. Li and S. Sra and S. Jegelka},
  booktitle={ICML},
  year={2016}
}
  • C. Li, S. Sra, S. Jegelka
  • Published in ICML 2016
  • Mathematics, Computer Science
  • We present a framework for accelerating a spectrum of machine learning algorithms that require computation of bilinear inverse forms uτ A-1 u, where A is a positive definite matrix and u a given vector. Our framework is built on Gausstype quadrature and easily scales to large, sparse matrices. Further, it allows retrospective computation of lower and upper bounds on uτ A-1 u, which in turn accelerates several algorithms. We prove that these bounds tighten iteratively and converge at a linear… CONTINUE READING
    12 Citations
    Faster Greedy MAP Inference for Determinantal Point Processes
    • 10
    • PDF
    Shifted Lanczos method for quadratic forms with Hermitian matrix resolvents
    Probability Density Rank-Based Quantization for Convex Universal Learning Machines
    • 3
    Polynomial time algorithms for dual volume sampling
    • 27
    • PDF
    Robust Guarantees of Stochastic Greedy Algorithms
    • 8
    • Highly Influenced
    • PDF
    Fast Mixing Markov Chains for Strongly Rayleigh Measures, DPPs, and Constrained Sampling
    • 26
    • PDF
    Fast DPP Sampling for Nystrom with Application to Kernel Methods
    • 53
    • PDF
    Fast Greedy MAP Inference for Determinantal Point Process to Improve Recommendation Diversity
    • 36
    • PDF
    DPPNet: Approximating Determinantal Point Processes with Deep Networks
    • 4
    • PDF

    References

    SHOWING 1-10 OF 74 REFERENCES
    Bounds for the Entries of Matrix Functions with Applications to Preconditioning
    • 103
    • PDF
    Some large-scale matrix computation problems
    • 186
    • Highly Influential
    • PDF
    Spectral methods in machine learning and new strategies for very large datasets
    • 135
    • PDF
    Revisiting the Nystrom Method for Improved Large-scale Machine Learning
    • 282
    • PDF
    Low cost high performance uncertainty quantification
    • 55
    • PDF
    Matrices, Moments and Quadrature with Applications
    • 206
    • PDF
    A probing method for computing the diagonal of a matrix inverse
    • J. Tang, Y. Saad
    • Mathematics, Computer Science
    • Numer. Linear Algebra Appl.
    • 2012
    • 97
    • Highly Influential
    • PDF
    An estimator for the diagonal of a matrix
    • 124
    • Highly Influential
    • PDF