Skip to main content
Theorem
Path
Curriculum
Paths
Labs
Diagnostic
Case Study
Blog
Search
Sign in
Quiz Hub
/
Trust Region Methods
Trust Region Methods
1 selected
Difficulty 5-5
1 unseen
View topic
Intermediate
New
0 answered
1 intermediate
Adapts to your performance
Question 1 of 1
120s
intermediate (5/10)
state theorem
Trust-region methods adjust the radius
Δ
k
based on the ratio
ρ
k
=
(
f
(
x
k
)
−
f
(
x
k
+
p
k
))
/
(
m
k
(
0
)
−
m
k
(
p
k
))
. What does
ρ
k
measure?
Hide and think first
A.
Step length divided by trust-region radius; controls whether the step hits the boundary or lies in the interior
B.
KL divergence between successive iterates' distributions; used in TRPO-style policy-gradient versions
C.
Gradient norm divided by Hessian norm; a Lipschitz-style bound that determines local convergence rate
D.
Actual decrease vs predicted decrease; close to 1 means the model is accurate,
<
0
means the step worsened
f
Submit Answer
I don't know