Linear Transformation Methods

Posted by JoSeK at 11:38 AM . Saturday, June 10, 2006
  • Agregar a Technorati
  • Agregar a Del.icio.us
  • Agregar a DiggIt!
  • Agregar a Yahoo!
  • Agregar a Google
  • Agregar a Meneame
  • Agregar a Furl
  • Agregar a Reddit
  • Agregar a Magnolia
  • Agregar a Blinklist
  • Agregar a Blogmarks

I'm interested on Linear Transformation Methods that allow us to transform an initial representation into another representation where the components are, in some way, independents. Our work with FBL (a Wrapper to improve Naive Bayes by deleting the dependent attributes) tries to do something similar and I would compare all those methods to FBL. The Linear Transformation Methods I've found are

  1. Principal Component Analysis
  2. Factor Analysis
  3. Projection Pursuit
  4. Independent Component Analysis
  5. Independent Factor Analysis
  6. Generalized Additive Models

0 comments:

Subscribe to: Post Comments (Atom)
Post a Comment