Articles
We consider the use of Gaussian process (GP) priors for solving inverse problems in a Bayesian framework. As is well known, the computational complexity of GPs scales cubically in the number of datapoints. Here we show that in the context of inverse problems involving integral operators, one faces a...