Beta. Content is under active construction and has not been peer-reviewed. Report errors on
GitHub
.
Disclaimer
Theorem
Path
Curriculum
Paths
Demos
Diagnostic
Search
Quiz Hub
/
Metropolis-Hastings Algorithm
Metropolis-Hastings Algorithm
5 questions
Difficulty 5-6
View topic
Intermediate
0 / 5
5 intermediate
Adapts to your performance
1 / 5
intermediate (5/10)
conceptual
Gibbs sampling draws each variable
X
i
from its full conditional
p
(
X
i
∣
X
−
i
)
in turn. When is Gibbs sampling especially appealing compared to Metropolis-Hastings?
Hide and think first
A.
When the joint distribution is hard to evaluate, so any sampler that avoids the joint altogether will perform better
B.
When the full conditionals
p
(
X
i
∣
X
−
i
)
have standard forms that can be sampled directly without rejection
C.
When the joint distribution is unimodal, so Gibbs always mixes faster than Metropolis regardless of parameter geometry
D.
When the posterior has heavy tails, because Gibbs sampling automatically handles tail behavior better than MH
Submit Answer