08:1408:14, 26 March 2014diffhist−19
Gibbs sampling
→Implementation: \mathbf{X} is multi-dimensional indeed, but if we say an initial value \mathbf{X}^{(0)} to each variable that is not correct since \mathbf{X} is itself just one variable; saying to each variable x_j instead would be right as well.
08:1408:14, 26 March 2014diffhist−19
Gibbs sampling
→Implementation: \mathbf{X} is multi-dimensional indeed, but if we say an initial value \mathbf{X}^{(0)} to each variable that is not correct since \mathbf{X} is itself just one variable; saying to each variable x_j instead would be right as well.