Using the Scipy implementation.
The NICTA library for Generalised Bayesian linear regression, and kernel-like models that scale to large data. See more at:
This is a NICTA developed module implementing Gaussian Process regression. This component is not documented in detail here, but its key components include:
Building and learning kernels:
Supervised prediction from a training dataset:
| condition(X, y, kernelFn, hyper_opt_config_copys) | |
| query(Xs, p) | |
| add_data(newX, newY, regressor[, query, ...]) | |
| mean(regressor, query) | |
| variance(regressor, query) | |
| predict |