Principal Component Analysis and finding important weights

From: "erdem" <salihoglu.erdem@xxxxxxxxx>
Date: 14 Aug 2006 12:19:45 -0700

Hi,

Suppose we have a classification problem at hand, and we want to find
the features that are important.

As I understood PCA is reducing the feature space, and transforming it.

Can we use PCA for finding important features of the problem.

Thanks

Erdem

.

Follow-Ups:
- Re: Principal Component Analysis and finding important weights
  - From: ImageAnalyst
- Re: Principal Component Analysis and finding important weights
  - From: Allan Lyckegaard

Prev by Date: Re: Image processing job questions..
Next by Date: Re: Principal Component Analysis and finding important weights
Previous by thread: 5x5 Kirsch operator kernel
Next by thread: Re: Principal Component Analysis and finding important weights
Index(es):
- Date
- Thread

Relevant Pages

Re: Principal Component Analysis and finding important weights
... how can we give weights to the features of ... and is it reasonable to use PCA based weightening for finding the ... feature space, i.e., it is already more or less orthogonal to the others. ...
(sci.image.processing)
Re: Principal Component Analysis and finding important weights
... PCA is a good mathematical technique for Dimensionality ... best way of finding the importance of the features.. ... Suppose we have a classification problem at hand, ...
(sci.image.processing)
Re: Principal Component Analysis and finding important weights
... best way of finding the importance of the features.. ... erdem wrote: ... Suppose we have a classification problem at hand, ... As I understood PCA is reducing the feature space, ...
(sci.image.processing)
Re: Principal Component Analysis and finding important weights
... erdem wrote: ... Suppose we have a classification problem at hand, ... As I understood PCA is reducing the feature space, ... Can we use PCA for finding important features of the problem? ...
(sci.image.processing)
Re: Principal Component Analysis and finding important weights
... Suppose we have a classification problem at hand, ... As I understood PCA is reducing the feature space, ... Can we use PCA for finding important features of the problem. ...
(sci.image.processing)