The Khintchine inequality , is a result in probability also frequently used in analysis bounding the expectation a weighted sum of Rademacher random variables with square-summable weights. It is named after Aleksandr Khinchin and spelled in multiple ways in the Latin alphabet.
It states that for each
p
∈
(
0
,
∞
)
{\displaystyle p\in (0,\infty )}
there exist constants
A
p
,
B
p
>
0
{\displaystyle A_{p},B_{p}>0}
depending only on
p
{\displaystyle p}
such that for every sequence
x
=
(
x
1
,
x
2
,
…
)
∈
ℓ
2
{\displaystyle x=(x_{1},x_{2},\dots )\in \ell ^{2}}
, and i.i.d. Rademacher random variables
ϵ
1
,
ϵ
2
,
…
{\displaystyle \epsilon _{1},\epsilon _{2},\dots }
,
A
p
≤
E
[
|
∑
n
=
1
∞
ϵ
n
x
n
|
p
]
1
/
p
‖
x
‖
2
≤
B
p
.
{\displaystyle A_{p}\leq {\frac {\mathbb {E} \left[\left|\sum _{n=1}^{\infty }\epsilon _{n}x_{n}\right|^{p}\right]^{1/p}}{\|x\|_{2}}}\leq B_{p}.}
As a particular case, consider
N
{\displaystyle N}
complex numbers
x
1
,
…
,
x
N
∈
C
{\displaystyle x_{1},\dots ,x_{N}\in \mathbb {C} }
, which can be pictured as vectors in a plane. Now sample
N
{\displaystyle N}
random signs
ϵ
1
,
…
,
ϵ
N
∈
{
−
1
,
+
1
}
{\displaystyle \epsilon _{1},\dots ,\epsilon _{N}\in \{-1,+1\}}
, with equal independent probability. The inequality states that
|
∑
i
ϵ
i
x
i
|
≈
|
x
1
|
2
+
⋯
+
|
x
N
|
2
{\displaystyle {\Big |}\sum _{i}\epsilon _{i}x_{i}{\Big |}\approx {\sqrt {|x_{1}|^{2}+\cdots +|x_{N}|^{2}}}}
with a bounded error.