CMLA

doc: gp_inf_update

Bayesian system update with new observations at Xt2

Syntax

 BayesInv = gp_inf_update(Ktt12, Ktt22, Yt, noise, BayInv1)
 BayesInv = gp_inf_update(Ktt12, Ktt22, Yt, noise, BayInv1, Ht)

Arguments

  • Ktt12 matrix (nt1, nt2) of kernel between the points of Xt1 and Xt2
  • Ktt22 matrix (nt2, nt2) of kernel between the points of Xt2
  • Yt vector (nt1+nt2, 1) of observations
  • noise noise standard deviation $\eta^2$
  • BayInv1 old system resolution for Xt1
  • Ht matrix (nt1+nt2, b) of basis data

Outputs

struct array containing:

  • RC upper triangular matrix (nt1+nt2,nt1+nt2) of Cholesky decomposition of Ktt+noise*I
  • invCY vector (nt1+nt2,1) solution of $(K + \eta^2 I)^{-1} Y$
  • beta vector (b,1) solution of the basis system

See also

gp_inf