NIPS2009_0092_slide - • Consider functions on graphs ,...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
Regression on Random Graphs Peter Sollich, Matthew Urry, Camille Coti GP regression : Bayesian method , predicts unknown function values We understand how well we can learn on continuous spaces – what about discrete ones?
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: • Consider functions on graphs , with random walk kernel • Focus on random graphs with fixed equal degree • Locally tree-like - use tree approximation? • Good for learning curves • Not the whole story for kernel shapes . . ....
View Full Document

Ask a homework question - tutors are online