Codex Wiki
OurBigBook.com
Site
Source code
Quadratic convergence bound for Newton's method
Home
Mathematics
Area of mathematics
Mathematical optimization
Newton method
OurBigBook.com
Words: 31
If
∇
2
f
⪰
m
I
and the Hessian is
M
-Lipschitz near a minimizer
x
∗
, then
∥
x
k
+
1
−
x
∗
∥
≤
2
m
M
∥
x
k
−
x
∗
∥
2
.
(2)
Thus a sufficiently close initial point has an error exponent that doubles at each iteration.
Ancestors
(5)
Newton method
Mathematical optimization
Area of mathematics
Mathematics
Home
Incoming links
(1)
Solution