Data-Driven Bandwidth Choice for Gamma Kernel Estimates of Density Derivatives on the Positive Semi-Axis.
Alexander V. DobrovidovLiubov MarkovichPublished in: ALCOSP (2013)
Keyphrases
- data driven
- density estimates
- density estimation
- higher order
- nonparametric density estimation
- positive and negative
- multiple instance learning
- kernel methods
- kernel function
- kernel regression
- kernel density estimator
- optimal kernel
- kernel matrix
- bandwidth allocation
- network bandwidth
- multiscale
- kernel density estimation
- kernel learning
- cross section
- probability density
- estimation error
- density function