Tailoring density estimation via reproducing kernel moment matching
2008pp. 992–999
Citations Over TimeTop 10% of 2008 papers
Abstract
Moment matching is a popular means of parametric density estimation. We extend this technique to nonparametric estimation of mixture models. Our approach works by embedding distributions into a reproducing kernel Hilbert space, and performing moment matching in that space. This allows us to tailor density estimators to a function class of interest (i.e., for which we would like to compute expectations). We show our density estimation approach is useful in applications such as message compression in graphical models, and image classification and retrieval.
Related Papers
- → Review Papers: Recent Developments in Nonparametric Density Estimation(1991)283 cited
- → Efficient Nonparametric Density Estimation on the Sphere with Applications in Fluid Mechanics(2000)22 cited
- → Variable Width Elliptic Gaussian Kernels for Probability Density Estimation(2007)5 cited
- Application of Nonparametric Weighted Kernel Density Estimation in Flood Frequency Analysis(2008)
- → Local Transformation Kernel Density Estimation of Loss Distributions(2009)