Next:
   Theory

   Contact:
   Peter Mills

Support This Project

libagf is a Peteysoft project

 A note to users of this site: If you've been following it, you know that I've recently started to include advertising. This is because no one has clicked on any of the donation buttons which are prominently displayed just about everywhere. Please take the time. Keep free software truly free. If all you can spare is $5, it may make all the difference.

Adaptive Gaussian Filtering


Adaptive Gaussian Filtering is a simple and powerful implementation of variable bandwidth kernel estimators for classification, PDF estimation and interpolation. The library incorporates several innovations to produce one of the fastest and most accurate supervised statistical classification algorithms in the world. These include:
  • matching kernel width to sample density quickly and accurately
  • restricting calculations to a set of k-nearest-neighbours found in O(n) time
  • generating a pre-trained model by searching for the class-borders with guaranteed, superlinear convergence
  • extrapolating the conditional probabilities to provide solid knowledge of estimate accuracy
For the latest information on this software, including updates, planned improvements and theoretical discussions, please check the Peteysoft homepage or the Notational Shorthand blog.

Author: Peter Mills

To cite this work:
Next: Theory