A library for gaussian process regression, I write this library for my Ph.D research of Bayesian optimization
- ARD Gaussian Kernel
- Constant mean function
- MLE to estimate the hyperparameters
- NLOPT
- MVMO
- VFE approximation
- Diagonal of covariance function
The training data should be stored in text files train_x and train_y, the test data should be written in test_x, see data/ for example
To use the first num_train data in the train_x and train_y as training set:
$ gp num_train
$ fitc num_train num_inducing