Evan Munro

Faster computation of matrix inverses and determinants with discrete x

1

@brad-ross and I are working on a paper that uses Stheno for GP related computation. If y = f(x) + e, f(x)~GP(m(x), k(x, x)), and x (known as the "running...

Cross-Validation (pseudo-likelihood) Criterion

1

I am finding for a specific use case for Gaussian Process Regression that optimizing kernel hyperparameters using the marginal likelihood `logpdf` results in poor generalization. I would like to try...

Integrate BayesRDD

2

@brad-ross and I have a currently private Julia module for RDD with Gaussian Process priors that is fairly organized and includes some nice speed-ups for discrete running variables. If there...