An algorithm shared by Ashok Srivastava, updated on Sep 10, 2010
The code in the stableGP package implements Gaussian process calculations using efficient and numerically stable algorithms. Description of the algorithms is in the paper "Stable and Efficient Gaussian Process Calculations" by L. Foster, A. Waagen, N. Aijaz, M. Hurley, A. Luis, J. Rinsky, C. Satyavolu, M. Way, P. Gazis, and A. Srivastava accepted in the Journal of Machine Learning Research, February, 2009.
The easiest way to get started using the code is to download and unzip the zip file, start Matlab (7.0 or higher), move to the appropriate folder and type either "demo_bootstrap" or "demo_history" to run one of the demonstration files.
Code in the zip file is based on the code used in the text Gaussian Processes or Machine Learning by Rasmussen and Williams (www.gaussianprocess.org/gpml/). So that the demonstrations are self contained and do not require that the user download additional code the zip file contains a few functions that are copied directly from Rasmussen and Williams' code.
|7.7 MB||91 downloads|
|291.5 KB||247 downloads|
For any questions, contact this resource's administrator: ashok22