Recomputes the likelihood parameter for exponential noise.
[newLik] = em_lapl(oldLikPar,y,cavM,cavV,postM,postV);
[newLik] = em_lapl(oldLikPar,y,cavM,cavV,postM,postV)
- the M-step in
lik.par. estimation of the coefficeient of the Laplace noise. The data is
assumed to be independent, thus an analytic expression is found (see MATLAB
code).
Parameters:
oldLikPar
- preceeding value of the likelihood parameter.
y
- training output.
cavM,cavV
- (cavity) mean and variance of the GP.
postM,postV
- mean and variance of the posterior process.
newLik
- the new parameter.
demogp_reg
, ogp
, ogptrain
, c_reg_lapl
Copyright (c) Lehel Csató (2001-2004)