Designed to be partially specified. (see examples)
SE(X, sigma = 1, rho = median(as.matrix(dist(t(X)))), jitter = 1e-10)LINEAR(X, sigma = 1, c = rep(0, nrow(X)))
Gram Matrix (N x N) (e.g., the Kernel evaluated at each pair of points)
covariate (dimension Q x N; i.e., covariates x samples)
scalar parameter
scalar bandwidth parameter
small scalar to add to off-diagonal of gram matrix (for numerical underflow issues)
vector parameter defining intercept for linear kernel
Gram matrix G is given by
SE (squared exponential): $$G = \sigma^2 * exp(-[(X-c)'(X-c)]/(s*\rho^2))$$
LINEAR: $$G = \sigma^2*(X-c)'(X-c)$$