Forum Freshman
Joined: 08 Jul 2008
Posts: 3
|
Hi,
I found a paper that discusses higher order moments for a 2d distribution. The central moments give the directions and length of the major and minor axis thus allowing me to draw an ellipse around the distribution.
How is this different from regular eigenvector calculations of the covariance matrix?
Also, how can I extend the concept to where it works for a 3d distribution thus allowing me to draw an ellipsoid?
The paper can be found at
http://ieeexplore.ieee.org/iel5/8352/26315/01167130.pdf?arnumber=1167130
Any ideas would be appreciated
thanks |
|
