The Sparse GP software is based on the following papers:

- L. Csato, E. Fokue, M. Opper, B. Schottky, O. Winther (2000):
Efficient Approaches to Gaussian Process Classification,
*Advances in Neural Information Processing Systems 13*. A draft version is at NCRG/99/025 - L. Csato, M. Opper (2002): Sparse Online Gaussian Processes,
*Neural Computation, 14/3 641-669*, draft version available as a draft report at NCRG/2001/014 - L. Csato (2002): Gaussian Processes - Iterative Sparse
Approximation,
*PhD Thesis, Aston University*, available from NCRG/2002/013 -
Here you can find the
*HTML version of the thesis*where most of the mathematical details are explained. - Here you can find the List of References from the thesis.

Some collections of references related to Gaussian Processes:

- By David J.C MacKay: http://www.inference.phy.cam.ac.uk/mackay/GP
- By Carl Rasmussen: www.cs.toronto.edu/~carl/gp.html
- By Mark Gibbs: http://www.inference.phy.cam.ac.uk/mng10/GP/

Since Gaussian Processes are probabilistic Kernel (Support Vector)
Machines, a source of information can also be the Kernel Machines web-site:
www.kernel-machines.org, regarding introductory material about
kernels.

References can also be found in the Examples section.

Questions, comments, suggestions: contact Lehel Csató.