[mlpack-svn] [MLPACK] #200: compute kernel convolution integrals
MLPACK Trac
trac at coffeetalk-1.cc.gatech.edu
Mon Feb 27 14:15:21 EST 2012
#200: compute kernel convolution integrals
------------------------------------------------------------------------+---
Reporter: nslagle | Owner: nslagle
Type: defect | Status: accepted
Priority: major | Milestone: mlpack 1.0.1
Component: mlpack | Resolution:
Keywords: kernel, convolution, least squares cross validation score | Blocking: 147, 148
Blocked By: 199 |
------------------------------------------------------------------------+---
Comment (by nslagle):
We'll be able to obtain volume expressions for the spherical, Gaussian,
and Epanechnikov kernels. (I'll look at other available kernels in the
kernel directory.) The convolution integral for the Gaussian is trivial
(simply another Gaussian kernel evaluation); the integral for the
spherical requires a numerical approximation of a high degree cosine
integral; the Epanechnikov will require more work, as the density
expression becomes quite messy over higher dimensions.
--
Ticket URL: <https://trac.research.cc.gatech.edu/fastlab/ticket/200#comment:3>
MLPACK <www.fast-lab.org>
MLPACK is an intuitive, fast, and scalable C++ machine learning library developed by the FASTLAB at Georgia Tech under Dr. Alex Gray.
More information about the mlpack-svn
mailing list