CWGP builds non-Gaussian GPs from compositions of elementary warpings with explicit inverses, delivering analytical predictions and improved accuracy and speed over standard WGPs.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Compositionally-Warped Gaussian Processes
CWGP builds non-Gaussian GPs from compositions of elementary warpings with explicit inverses, delivering analytical predictions and improved accuracy and speed over standard WGPs.