Preconditioned Krylov solvers for kernel regression

@article{Srinivasan2013PreconditionedKS,
  title={Preconditioned Krylov solvers for kernel regression},
  author={Balaji Vasan Srinivasan and Qi Hu and Nail A. Gumerov and Raghu Murtugudde and Ramani Duraiswami},
  journal={CoRR},
  year={2013},
  volume={abs/1408.1237}
}
A primary computational problem in kernel regression is solution of a dense linear system with the N × N kernel matrix. Because a direct solution has an O(N) cost, iterative Krylov methods are often used with fast matrix-vector products. For poorly conditioned problems, convergence of the iteration is slow and preconditioning becomes necessary. We investigate preconditioning from the viewpoint of scalability and efficiency. The problems that conventional preconditioners face when applied to… CONTINUE READING
3 Citations
25 References
Similar Papers

References

Publications referenced by this paper.
Showing 1-10 of 25 references

GPUML: Graphi cal processors for speeding up kernel machines

  • B. Srinivasan, Q. Hu, R. Duraiswami
  • Workshop on High Performance Analytics…
  • 2010
Highly Influential
7 Excerpts

The improved fast Gauss tra nsform with applications to machine learning

  • V. Raykar, R. Duraiswami
  • Large Scale Kernel Machines , 2007, pp. 175–201.
  • 2007
3 Excerpts

Sparse Gaussian process es u ing pseudo-inputs

  • E. Snelson, Z. Ghahramani
  • inAdvances in Neural Information Processing…
  • 2006
3 Excerpts

Similar Papers

Loading similar papers…