David MacKay

Environment and Energy 
Information Theory 
·Gallager codes & MN codes 
·Theory of Sparse Graph codes 
·Regular Gallager codes 
·Irregular Gallager codes 
·Turbo codes  
·Insertions and deletions 
·Quantum Codes 
·Cryptanalysis and Coding 
·Discussion documents 
·More Message-Passing 
Probabilistic Data Modelling 
·PhD thesis 
·Bayes for neural networks 
·ICA & Latent Variable Models 
·Bayes FAQ 
·Bayesian computations « 
·Monte Carlo methods 
·Variational methods 
·Gaussian processes 
·Language, Proteins, Density nets 
·Images, Radar, Neuroscience 
·Materials science 
·Decision theory, Fair Elections 
Human-computer Interfaces 
Computational Neuroscience 
Evolution and Molecular Biology  
Miscellaneous work 
Software and Files 
Any Problems? 

Search :


Computational Implementation of Bayesian methods

alpha.ps.gz, | PDF | DJVU | abstract | ps mirror, Canada | pdf | djvu |
Hyperparameters: optimize, or integrate out?
This paper appeared in shortened form in the 1993 Maxent proceedings. The long version underwent a lengthy refereeing process with Neural Computation, where it was published (1999) under the title Comparison of Approximate Methods for Handling Hyperparameters.
laplace.ps.gz. | PDF | DJVU | abstract. | <- UK | -> Canada | laplace.ps.gz. abstract.
`Choice of Basis for Laplace Approximation'.
Published in Machine Learning. Volume 33, No. 1, October 1998
(Submitted October 14th 1996)
delve.ps.gz. | PDF | DJVU | abstract. | <- UK | Canada -> | delve.ps.gz. abstract.
` More sensitive tests in DELVE'.

See also

The Inference Group is supported by the Gatsby Foundation
and by a partnership award from IBM Zurich Research Laboratory
David J.C. MacKay
Site last modified Wed Nov 12 17:38:44 GMT 2014