Skip to content
/ sigp Public

Finite-Sample Integral Gaussian Processes

Notifications You must be signed in to change notification settings

ztanml/sigp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Finite-Sample Integral Gaussian Processes (SIGP)

  • A dual construction of Gaussian processes with sample paths in a given reproducing kernel Hilbert space.
  • Matlab code for the paper: https://arxiv.org/abs/1802.07528

SIGP on Real-Life Data

Example 1: Classification of the Arcene cancer data

Data source: https://archive.ics.uci.edu/ml/datasets/Arcene

In Matlab:

>> demoSIGPArcene
Loading the data ...
Classifying with SIGP ...
F1 score:0.85714

For comparison, the standard GP based on GPML Toolbox (http://www.gaussianprocess.org/gpml/code/matlab/doc/) yields a lower F1 score 0.82353. To verify, add GPML to the Matlab PATH, and run demoGPArcene.m.

Example 2: Prediction on Boston housing data

Data source: https://archive.ics.uci.edu/ml/machine-learning-databases/housing/

In Matlab:

>> demoSIGPHousing
Loading housing data ...
Training SIGP ...
Mean squared error:28.1999

For comparison, the standard GP based on GPML Toolbox (http://www.gaussianprocess.org/gpml/code/matlab/doc/) yields a much larger mean squared error 93.1109. To verify, add GPML to the Matlab PATH, and run demoGPHousing.m.

About

Finite-Sample Integral Gaussian Processes

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages