h******3
发帖数: 190 | 1 Is there any theory that says the algorithm will eventually converge to the
right distribution regardless of the starting values?
thanks~ | m********9
发帖数: 208 | | h******3
发帖数: 190 | 3 谢谢你. 我其实只想知道是不是最终都可以converge到true distribution regardless
of starting values.
【在 m********9 的大作中提到】
: SLLN
| g*******i
发帖数: 258 | 4 it should be. This is the idea of Gelman-Rubin convergence diagnostic
if the markov chain has good mixing, then no matter where you start, it will
converge to a same level
regardless
【在 h******3 的大作中提到】
: 谢谢你. 我其实只想知道是不是最终都可以converge到true distribution regardless
: of starting values.
|
|
