Gaussian quadrature for matrix inverse forms with applications

@inproceedings{Li2016GaussianQF,
  title={Gaussian quadrature for matrix inverse forms with applications},
  author={Chengtao Li and Suvrit Sra and Stefanie Jegelka},
  booktitle={ICML},
  year={2016}
}
We present a framework for accelerating a spectrum of machine learning algorithms that require computation of bilinear inverse forms u>A 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 u>A u, which in turn accelerates several algorithms. We prove that these bounds tighten iteratively and converge at a linear… CONTINUE READING

Similar Papers

Loading similar papers…