Corpus ID: 16765297

A Linear Approximation to the chi^2 Kernel with Geometric Convergence

@article{Li2012ALA,
  title={A Linear Approximation to the chi^2 Kernel with Geometric Convergence},
  author={F. Li and G. Lebanon and C. Sminchisescu},
  journal={arXiv: Learning},
  year={2012}
}
  • F. Li, G. Lebanon, C. Sminchisescu
  • Published 2012
  • Mathematics, Computer Science
  • arXiv: Learning
  • We propose a new analytical approximation to the $\chi^2$ kernel that converges geometrically. The analytical approximation is derived with elementary methods and adapts to the input distribution for optimal convergence rate. Experiments show the new approximation leads to improved performance in image classification and semantic segmentation tasks using a random Fourier feature approximation of the $\exp-\chi^2$ kernel. Besides, out-of-core principal component analysis (PCA) methods are… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 43 REFERENCES
    Efficient Additive Kernels via Explicit Feature Maps
    619
    Random Fourier Approximations for Skewed Multiplicative Histogram Kernels
    68
    Nyström Method vs Random Fourier Features: A Theoretical and Empirical Comparison
    230
    On the Nyström Method for Approximating a Gram Matrix for Improved Kernel-Based Learning
    761
    Generalized RBF feature maps for Efficient Detection
    81
    Efficient Classification for Additive Kernel SVMs
    199
    Using the Nyström Method to Speed Up Kernel Machines
    1762
    Efficient SVM Training Using Low-Rank Kernel Representations
    632
    The Quadratic-Chi Histogram Distance Family
    238
    Locality-constrained Linear Coding for image classification
    3075