W. Lio and B. Liu / Residual and confidence interval for uncertain regression model 2575
As a real-valued function on the uncertainty space
(, L, M), uncertain variable is introduced to model
the quantity with human uncertainty.
Definition 2.2. (Liu [13]) An uncertain variable ξ
is a measurable function from an uncertainty space
(, L, M) to the set of real numbers such that for any
Borel set B of real numbers, the set
{ξ ∈ B}={γ ∈ | ξ(γ) ∈ B}
is an event.
The uncertainty distribution of an uncertain vari-
able ξ is defined by (x) = M{ξ ≤ x} for any real
number x. An uncertainty distribution (x) is said to
be regular if it is a continuous and strictly increasing
function with respect to x at which 0 <(x) < 1,
and
lim
x→−∞
(x) = 0, lim
x→∞
(x) = 1.
If ξ is an uncertain variable with regular uncer-
tainty distribution (x), the inverse function
−1
(α)
is called the inverse uncertainty distribution of ξ (Liu
[15]).
An uncertain variable ξ is called linear if it has an
uncertainty distribution
(x) =
⎧
⎪
⎨
⎪
⎩
0, if x ≤ a
(x − a)/(b − a), if a<x≤ b
1, if x>b
denoted by L(a, b), where a and b are real numbers
satisfying a<b, and the inverse uncertainty distri-
bution of linear uncertain variable L(a, b)is
−1
(α) = (1 − α)a + αb.
An uncertain variable ξ is called zigzag if it has an
uncertainty distribution
(x) =
⎧
⎪
⎪
⎪
⎨
⎪
⎪
⎪
⎩
0, if x ≤ a
(x − a)/[2(b − a)], if a<x≤ b
(x + c − 2b)/[2(c − b)], if b<x≤ c
1, if x>c
denoted by Z(a, b, c), where a, b and c are real num-
bers satisfying a<b<c, and the inverse uncertainty
distribution of zigzag uncertain variable Z(a, b, c)is
−1
(α) =
(1 − 2α)a + 2αb, if α<0.5
(2 −2α)b + (2α − 1)c, if α ≥ 0.5.
An uncertain variable ξ is called normal if it has an
uncertainty distribution
(x) =
1 + exp
π(e − x)
√
3σ
−1
,x ∈
denoted by N(e, σ), where e and σ are real numbers
satisfying σ>0, and the inverse uncertainty distri-
bution of normal uncertain variable N(e, σ)is
−1
(α) = e +
σ
√
3
π
ln
α
1 − α
.
Definition 2.3. (Liu [14]) The uncertain variables ξ
1
,
ξ
2
, ···, ξ
n
are said to be independent if
M
n
i=1
(ξ
i
∈ B
i
)
=
n
i=1
M
{
ξ
i
∈ B
i
}
for any Borel sets B
1
, B
2
, ···, B
n
of real numbers.
Assume that ξ
1
,ξ
2
, ···,ξ
n
are independent uncer-
tain variables with regular uncertainty distributions
1
,
2
, ···,
n
, respectively. Liu [15] showed that
if f (x
1
,x
2
, ···,x
n
) is a strictly monotonous func-
tion, then the inverse uncertainty distribution of the
uncertain variable f (ξ
1
,ξ
2
, ···,ξ
n
) can be calcu-
lated by the following theorems.
Theorem 2.1. (Liu [15]) Let ξ
1
,ξ
2
, ···,ξ
n
be
independent uncertain variables with regular
uncertainty distributions
1
,
2
, ···,
n
, respec-
tively. If f is strictly increasing with respect to
ξ
1
,ξ
2
, ···,ξ
m
and strictly decreasing with respect
to ξ
m+1
,ξ
m+2
, ···,ξ
n
, then ξ = f (ξ
1
,ξ
2
, ···,ξ
n
)
is an uncertain variable with inverse uncertainty
distribution
−1
(α) = f (
−1
1
(α), ···,
−1
m
(α),
−1
m+1
(1 − α), ···,
−1
n
(1 − α)).
As the average value of an uncertain variable in
the sense of uncertain measure, expected value can
represent the size of the uncertain variable.
Definition 2.4. (Liu [13]) Let ξ be an uncertain vari-
able. Then the expected value of ξ is defined as
E[ξ] =
+∞
0
M{ξ ≥ x}dx −
0
−∞
M{ξ ≤ x}dx
provided that at least one of the two integrals is finite.
As another important feature for an uncertain vari-
able, variance is defined as follows: