Displaying 1 result from an estimated 1 matches for "inequalty".
Did you mean:
inequality
2008 Dec 08
1
Multivariate kernel density estimation
...esidual variance (s2). The means are
independent of eachother, but conditional on the residual variance. So now I
have a data frame of say 10.000 iterations, and k+1 parameters.
I am especially interested in the posterior distribution of the mu
parameters, because I want to test the support for an inequalty constrained
model (e.g. mu1 > mu2 > mu3). I wish to derive the multivariate 95% highest
density parameter space for the mu parameters. For example, if I had a
posterior distirbution with 2 means, this should somehow result in the
circle or elipse that contains the 95% highest density area....