10

HyperUncertain, SuperUncertain,
and SuperHyperUncertain
Sets/Logics/Probabilities/Statistics
Florentin Smarandache 1

1 University of New Mexico, Gallup, NM, USA

[email protected]

Abstract
In this paper, we define for the first time the HyperUncertain, SuperUncertain, and
SuperHyperUncertain Sets/Logics/Probabilities/Statistics, the classical case (when the
appurtenance degree of a generic element belongs to the unit interval [0, 1]), and the non-
classical case (when the appurtenance degree of a generic element is outside on the interval
[0, 1]; such theories are called Over/Under/Off-Sets/Logics/Probabilities/Statistics).
We prove that the Unary SuperHyperFunction is a generalization of all Uncertain
Sets/Logics/Probabilities.

Keywords
SuperHyperStructure, Unary SuperHyperFunction, (h-ary, k-ary) SuperHyperFunction,
HyperUncertain Sets, HyperUncertain Logics, HyperUncertain Probabilities, HyperUncertain
Statistics, SuperUncertain Sets, SuperUncertain Logics, SuperUncertain Probabilities,
SuperUncertain Statistics, SuperHyperUncertain Sets, SuperHyperUncertain Logics,
SuperHyperUncertain Probabilities, SuperHyperUncertain Statistics.

1. Denominations of “Uncertain” and “Theory”

By “Uncertain” we mean the following types: Fuzzy, Intuitionistic Fuzzy,

Neutrosophic, Inconsistent Intuitionistic Fuzzy (Picture Fuzzy, Ternary Fuzzy),
Pythagorean Fuzzy (Atanassov's Intuitionistic Fuzzy of second type), Fermatean
Fuzzy, q-Rung Orthopair Fuzzy, Spherical Fuzzy, n-HyperSpherical Fuzzy, Refined
Neutrosophic, Plithogenic, etc.

Critical Review. Volume XIV, 2017

11 Florentin Smarandache
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

By "Theory" we mean the following: Set, Logic, Probability, and Statistics.

2. Definition of SuperHyperStructure

A SuperHyperStructure [4] is a structure built on the n-th PowerSet of a Set A, for n ≥

1, as in our real world
{because a set (or system) A (that may be a set of items, an organization, a country,
etc.) is composed by sub-sets that are parts of P(A), which in their turn are organized
in sub-sub-sets that are parts of P(P(A)) = P2(A), then in sub-sub-sub-sets that are
parts of P3(A), and so on, Pn(A) = P(Pn-1(A) }.
As part of this most general notion, called SuperHyperStructure or (m, n)-
SuperHyperStructure, and its associated terms such as SuperHyperAxiom,
SuperHyperOperator, SuperHyperAlgebra etc. all introduced by Smarandache [1] in
2016, let’s recall the definitions of the n-PowerSet of a Set, and of the
SuperHyperFunctions that will be used in this paper.

3. Definition [1, 4] of the n-th PowerSet of a Set A

Let 𝒰 be a universe of discourse, and A be a non-empty subset of 𝒰,

with 𝑃(𝐴) 𝑡ℎ𝑒 𝑝𝑜𝑤𝑒𝑟𝑠𝑒𝑡 𝑜𝑓 𝐴, and n  0 an integer.
Then:
def
P 0 ( A) = A
P1 ( A) = P( A)
P 2 ( A) = P( P( A))
...............................
P n ( A) = P( P n −1 ( A))

4. Definition of the (h, k)-SuperHyperFunction

The most general form of a function is the (h-ary, k-ary)-SuperHyperFunction:

𝑓𝑆𝐻
𝑆𝐻
: 𝑃𝑚1 (𝐴1 ) × 𝑃𝑚2 (𝐴2 ) × … × 𝑃𝑚ℎ (𝐴ℎ ) → 𝑃𝑛1 (𝐵1 ) × 𝑃𝑛2 (𝐵2 ) × … × 𝑃𝑛𝑘 (𝐵𝑘 )
where 𝐴1 , 𝐴2 , …, 𝐴ℎ , 𝐵1 , 𝐵2 , …, 𝐵𝑘 are subsets of 𝒰, and h, k are integers ≥ 0.
For each x1  P m1 ( A1 ), x2  P m2 ( A2 ),..., xh  P mh ( Ah ) one has
SH
f SH ( x1 , x2 ,..., xh )  P n1 ( B1 )  P n2 ( B2 )  ...  P nk ( Bk ) ,
where “SH” on the bottom of f SHSH stands for “SuperHyper” (which means that one has
mi-powersets, 1  i  h ) in the domain of the function, while “SH” on the top of f SHSH
stands for “SuperHyper” (which similarly means that one has nj-powersets, 1  j  k )

Critical Review. Volume XIV, 2017

 Florentin Smarandache 12
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

in the range of the function.

5. Unary SuperHyperFunction

The Unary SuperHyperFunction is a particular case of the (h-ary, k-ary)-

SuperHyperFunction when h = k =1.
Let 𝒰 be a universe of discourse, and 𝐴, 𝐵 two non-empty subsets of 𝒰.
Let 𝑚, 𝑛 ≥ 0 be integers,
and P m ( A) be the m-th powerset of A, while similarly, P n ( B) be the n-th powerset of B.
Then: 𝑓𝑆𝐻
𝑆𝐻
: 𝑃𝑚 (𝐴) → 𝑃𝑛 (𝐵)
is called Unary SuperHyperFunction.
For an x  Pm ( A) one has an f SH
SH
( x)  P n ( B) .

6. Theorem:
The Unary SuperHyperFunction is a generalization of all classical and non-classical
Uncertain Sets/Logics/Probabilities/Statistics.
By “classical” we mean those theories whose codomain is the unit interval,
B = [0, 1].
By “non-classical” we mean the theories whose codomain is different from [0, 1], i.e.
B = [, ]  [0,1] , where   0  1   ,
as in Over/Under/Off-Sets/Logics/Probability/Statistics {Smarandache, [2, 3]}.

Proof:

Let 𝒰 be a universe of discourse, and 𝐴 a non-empty subset of 𝒰.

Let A be the domain of an Uncertain Set/Logic/Probability/Statistics, where the Uncertain Statistics
is the statistical characterization of its corresponding Uncertain Probability, therefore it may have
the same domain as its Uncertain Probability – no matter if one has a classical or non-classical theory
(see above).

Let B = [0, 1] be the real unit interval, as codomain.

(a) Then, as a particular case of the Unitary SuperHyperFunction, one considers the function:
SH
f SH : P m ( A) → P n ([0,1]) (*)
This function can easily represent any type of classical HyperUncertain SuperUncertain, and
SuperHyperUncertain Set/Logic/Probability/Statistics.
(b) While the function:
SH
f SH : P m ( A) → P n ([ , ]) , where   0  1   ,
represents any type of non-classical { i.e. Over/Under/Off- } HyperUncertain, SuperUncertain, and

Critical Review. Volume XIV, 2017

13 Florentin Smarandache
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

SuperHyperUncertain Set/Logic/Probability/Statistics [2, 3].

7. Number of Degree-Types

Let τ: A → N, where A is the domain of a given Uncertain Theory, while N is the set of natural
(positive integer) numbers, with 𝜏(𝑥) = the number of degree types of this given Uncertain Theory
(i.e. Uncertain Set/Logic/Probability/Statistics).

Let 𝜏(𝑥) = r ≥ 1 be the number of degree types ( d1, d2, …, dr ) of some Uncertain Theory,
for example, if r = 4 one may have:
d1 = degree of membership/truth/chance_of_occuring,
d2 = degree of nonmembership/falsehood/chance_of_not_occuring
d3 = degree of indeterminacy/neutrality
d4 = degree of contradiction.
Then:
d1 , d 2 ,..., dr : P m ( A) → P n ([0,1]) , for classical Uncertain Theories.
Or:
d1 , d 2 ,..., dr : P m ( A) → P n ([ , ]) ,   0  1   , for un-classical Uncertain Theories.
As theories, we especially refer to "Set" and "Logic" and “Probability” (since “Statistics” is
just characterization of the Probability events), which are the most used ones with respect
to uncertain(ty).

(i) For several Uncertain Theories, for any x  A, 𝜏(𝑥) is a single degree – degree of
membership (or truth, or chance of occurring), normally denoted by 𝑡(𝑥),
therefore 𝜏(𝑥) = 𝑡(𝑥),
or 𝜏: 𝐴 → [0,1], especially in classical fuzzy set/logic/probability.

(ii) For other Uncertain Theories, 𝜏(𝑥) is a double degree: degree of membership (or
truth), denoted by 𝑡(𝑥), and degree of nonmembership (or falsehood), denoted by 𝑓(𝑥).
Therefore: 𝜏(𝑥) = (𝑡(𝑥), 𝑓(𝑥)), or 𝜏: 𝐴 → [0,1]2 .
As an example: for intuitionistic fuzzy set/logic/probability and its derivatives.

(iii) For Neutrosophic Set/Logic/Probability and some of its derivatives, τ(x) is a triple
degree:
• degree of membership (or truth), 𝑡(𝑥);
• degree of indeterminacy (or neutrality), 𝑖(𝑥);
• and degree of nonmembership (or falsehood), 𝑓(𝑥).
As such, 𝜏(𝑥) = (𝑡(𝑥), 𝑖(𝑥), 𝑓(𝑥)), or 𝜏: 𝐴 → [0,1]3 .

Critical Review. Volume XIV, 2017

 Florentin Smarandache 14
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

(iv) For Refined Neutrosophic Set/Logic/Probability, and its derivatives, τ(x) is an q-tuple
degree ( q  1 ):
𝜏(𝑥) = (𝑡₁(𝑥), 𝑡₂(𝑥), . . . , 𝑡𝑝 (𝑥); 𝑖₁(𝑥), 𝑖₂(𝑥), . . . , 𝑖𝑟 (𝑥); 𝑓₁(𝑥), 𝑓₂(𝑥), . . . , 𝑓𝑠 (𝑥))
where 𝑡𝑗 (𝑥), 1 ≤ 𝑗 ≤ 𝑝, is type-𝑗 of membership (truth);
𝑖𝑘 (𝑥), 1 ≤ 𝑘 ≤ 𝑟, is type-𝑘 of indeterminacy (neutrality);
𝑓𝑙 (𝑥), 1 ≤ 𝑙 ≤ 𝑠, is type-𝑙 of nonmembership (falsehood),
with 𝑝 + 𝑟 + 𝑠 = 𝑞,
or 𝜏: 𝐴 → [0,1]𝑞 .

8. HyperUncertain Classical Set/Logic/Probability/Statistics

 : A → Pn ([0,1]r )

9. SuperUncertain Classical Set/Logic/Probability/Statistics

 : Pm ( A) → [0,1]r

10. SuperHyperUncertain Classical Set/Logic/Probability/Statistics

 : Pm ( A) → Pn ([0,1]r )
*
For non-classical Uncertain Theories, the unit interval [0, 1] is replaced by the non-unit interval
[ , ], where   0  1   .

11. HyperUncertain Non-Classical Set/Logic/Probability/Statistics

 : A → Pn ([ , ]r )

12. SuperUncertain Non-Classical Set/Logic/Probability/Statistics

 : Pm ( A) → [, ]r

13. SuperHyperUncertain Non-Classical Set/Logic/Probability/Statistics

 : Pm ( A) → Pn ([, ]r )

14. Examples of HyperUncertain Classical Set/Logic/Probability/Statistics

 : A → Pn ([0,1]r )

Critical Review. Volume XIV, 2017

15 Florentin Smarandache
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

"Hyper" means the degree of an element 𝑥 ∈ 𝐴 is a subset (not a single element) of

[0, 1].

(i) HyperFuzzy is equivalent

to Subset-Valued Fuzzy Set
Let 𝑡: 𝐴 → 𝑃([0,1]),
where for any x  A , t(x) represents the degree of
membership/truth/chance_of_occuring of the element x with respect to the set A.
t(x) is a subset (not a single element) of [0,1].

- The Interval-Valued Fuzzy is part of Hyperfuzzy, for examples:

𝑡(𝑎₁) = [0.2, 0.3], 𝑡(𝑎₂) = (0.4, 0.6).
- The Subset-Valued Fuzzy is also part of Hyperfuzzy, for example
𝑡(𝑎₁) = (0.4, 0.5) ∪ [0.6, 0.8].
- The Hesitant Fuzzy is also a part of Hyper Fuzzy, for example
t(𝑎₁) = {0.4, 0.7, 0.9}.

(ii) HyperIntuitionistic Fuzzy Set is equivalent

to Subset-Valued Intuitionistic Fuzzy Set
Let 𝒰 be a universe of discourse and A a non-empty subset of 𝒰.
𝑡: 𝐴 → 𝑃([0,1]) is the truth/membership function,
while 𝑓: 𝐴 → 𝑃([0,1]) is the falsehood/nonmembership function.
As particular cases of the HyperIntuitionistic Fuzzy Set, one has:

- Interval-Valued Intuitionistic Fuzzy Set

where 𝑡, 𝑓: 𝐴 → 𝑃([0,1]) are intervals, for example:

𝑡(𝑎1 ) = [0.1,0.4],
𝑓(𝑎1 ) = [0.2,0.5].

- Hesitant Intuitionistic Fuzzy Set

For example:
𝑡(𝑎1 ) = {0.2, 0.3, 0.4},
𝑓(𝑎1 ) = {0.1, 0.7}.

Critical Review. Volume XIV, 2017

 Florentin Smarandache 16
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

(iii) HyperNeutrosophic Set is equivalent to Subset-Valued Neutrosophic Set

Let 𝒰 be a universe of discourse and 𝐴 a non-empty subset of 𝒰.

𝑡, 𝑖, 𝑓: 𝐴 → 𝑃([0,1]),
where 𝑡(𝑎), 𝑖(𝑎), 𝑓(𝑎) are subsets in [0,1], for all 𝑎 ∈ 𝐴.
As particular cases one has:

- Subset-Valued Neutrosophic Set

When 𝑡(𝑎), 𝑖(𝑎), 𝑓(𝑎) are any types of subsets of [0, 1]. For example:
𝑡(𝑎1 ) = [01. , 0.2] ∪ (0.5, 0.6),
𝑖(𝑎1 ) = {0.6, 0.7, 0.9},
𝑓(𝑎1 ) = [03. , 0.4] ∪ {0.50, 0.51, 0.52}.
This was the general definition [1995] of the Standard Neutrosophic Set. (The Non-
Standard Neutrosophic Set is not the subject of this paper.)

- Interval-Valued Neutrosophic Set

where 𝑡(𝑎), 𝑖(𝑎), 𝑓(𝑎) are intervals included in [0,1], for all 𝑎 ∈ 𝐴.
For example: 𝑡(𝑎₁) = [0.6, 0.7], 𝑖(𝑎₁) = [0.5, 0.6], 𝑓(𝑎₁) = [0.8, 1.0].

- Hesitant-Valued Neutrosophic Set

When 𝑡(𝑎), 𝑖(𝑎), 𝑓(𝑎) are hesitant discrete subsets of [0, 1]. For example:
𝑡(𝑎1 ) = {0.2, 0.3, 0.4},
𝑖(𝑎1 ) = {0.70, 0.75},
𝑓(𝑎1 ) = {0.5, 0.6, 0.7, 0.8}.

(iv) HyperRefinedNeutrosophic Set is equivalent to Subset-Valued Refined

Neutrosophic Set

- Interval-Valued (2,3,1)-Refined Neutrosophic Set/Logic/Probability

is a Hyper (2,3,1)-Refined Neutrosophic Set/Logic/Probability
𝜏: 𝐴 → 𝑃([0,1]6 ), because 2 + 3 + 1 = 6.
𝜏(𝑎1 ) = (𝑡1 (𝑎1 ), 𝑡 2 (𝑎1 ); 𝑖 1 (𝑎1 ), 𝑖 2 (𝑎1 ), 𝑖 3 (𝑎1 ); 𝑓 1 (𝑎1 )) =
[0.1, 0.2], [0.3, 0.4]; (0.2, 0.3]; (0.2, 0.4), [0.3, 0.5); (0.7, 0.9]),
for example.

Critical Review. Volume XIV, 2017

17 Florentin Smarandache
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

15. Examples of SuperUncertain Classical Set/Logic/Probability/Statistics

 : Pm ( A) → [0,1]r
“Super” means that one considers the degree of a subset (not of a single element) of A.

- SuperFuzzy Set/Logic/Probability

τ: P(A) → [0,1]
but τ({a₁, a₂}) = t({a₁, a₂}) = 0.9, for example, which means that the membership/truth
degree of the whole subset {a₁, a₂} together being as a team is 0.9.

- SuperIntuitionistic Fuzzy Set/Logic/Probability

𝜏: 𝑃(𝐴) → [0,1]2 )
τ({𝑎1 , 𝑎2 }) = {t({𝑎1 , 𝑎2 }), f ({𝑎1 , 𝑎2 })} = { 0.5, 0.7 }, for example.
Or, t({𝑎1 , 𝑎2 }) = 0.5, and f({𝑎1 , 𝑎2 }) = 0.7.

- SuperNeutrosophic Set/Logic/Probability

𝜏: 𝑃(𝐴) → [0, 1]3

𝜏({𝑎1 , 𝑎2 }) = { 𝑡({𝑎1 , 𝑎2 }), 𝑖({𝑎1 , 𝑎2 }), 𝑓({𝑎1 , 𝑎2 }) } = { 0.8, 0.1, 0.3 }, for example.
Or, t({𝑎1 , 𝑎2 }) = 0.8, i({𝑎1 , 𝑎2 )} = 0.1, f({𝑎1 , 𝑎2 }) = 0.3.

16. Examples of SuperHyperUncertain Classical Set/Logic/Probability/Statistics

 : Pm ( A) → Pn ([0,1]r )
“Super” means that one considers the degree of a subset (not of a single element) of A.
While “Hyper” means that that the degree is a subset (not of a single element) of [0, 1].

- SuperHyperFuzzy Set/Logic/Probability

τ: P(A) → P([0,1])
but τ({a₁, a₂}) = t({a₁, a₂}) = [0.7, 0.8], for example.

- SuperHyperIntuitionistic Fuzzy Set/Logic/Probability

𝜏: 𝑃(𝐴) → 𝑃([0,1]2 )

Critical Review. Volume XIV, 2017

 Florentin Smarandache 18
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

τ({𝑎1 , 𝑎2 }) = {t({𝑎1 , 𝑎2 }), f ({𝑎1 , 𝑎2 })} = { [0.7, 0.8], [0.9, 1.0] }, for example.

- SuperHyperNeutrosophic Set/Logic/Probability

𝜏: 𝑃(𝐴) → 𝑃([0, 1]3 )

𝜏({𝑎1 , 𝑎2 }) = { 𝑡({𝑎1 , 𝑎2 }), 𝑖({𝑎1 , 𝑎2 }), 𝑓({𝑎1 , 𝑎2 }) } = {[0.7,0.8], [0.0,0.2], [0.9,1.0]},
for example.

17. HyperUncertain, SuperUncertain, and SuperHyperUncertain

Over/Under/Off-Sets/Logics/Probability/Statistics

Let A be an uncertain (set/logic/probability) domain, and B = [, ] be a real interval,

where   0  1   .
Then, as a particular case, one considers the function: f SH
SH
: P m ( A) → P n ([ , ]) ,
which similarly represent the HyperUncertain and SuperHyperUncertain
Over/Under/Off-Sets/Logics/Probability/Statistics.

18. Examples of HyperUncertain, SuperUncertain, and SuperHyperUncertain

Over/Under/Off-Sets/Logics/Probability/Statistics

- HyperFuzzy Non-Classical Set/Logic/Probability

𝜏: 𝐴 → 𝑃([−0.2, 1])
𝜏(𝑎₁) = 𝑡(𝑎₁) = [−0.1, 0.3], for example.

- SuperFuzzy Non-Classical Set/Logic/Probability

𝜏: 𝑃(𝐴) → 𝑃([0, 1.1])

𝜏({𝑎1 , 𝑎2 }) = 𝑡({𝑎1 , 𝑎2 }) = 1.1, for example.

- SuperHyperFuzzy Non-Classical Set/Logic/Probability

τ: P(A) → P([-0.3, 1.2])

but τ({a₁, a₂}) = t({a₁, a₂}) = [0.9, 1.1], for example.

19. Conclusion
We have defined all Classical (and Non-Classical) HyperUncertain, SuperUncertain, and

Critical Review. Volume XIV, 2017

19 Florentin Smarandache
HyperUncertain, SuperUncertain, and SuperHyperUncertain Sets/Logics/Probabilities/Statistics

SuperHyperUncertain Set/Logic/Probability/Statistics. We also presented many

numerical examples that occur in our everyday world.

References

[1] F. Smarandache, SuperHyperAlgebra and Neutrosophic SuperHyperAlgebra,

Section into the author’s book: “Nidus Idearum. Scilogs, II: de rerum consectatione,”
Second Edition, (2016), pp. 107– 108, https://fs.unm.edu/NidusIdearum2-ed2.pdf
[2] Florentin Smarandache, Neutrosophic Overset, Neutrosophic Underset, and
Neutrosophic Offset Similarly for Neutrosophic Over-/Under-/OffLogic, Probability, and
Statistics, Pons Editions, Brussels, 2016,
https://fs.unm.edu/NeutrosophicOversetUndersetOffset.pdf
[3] F. Smarandache, Operators on Single-Valued Neutrosophic Oversets, Neutrosophic
Undersets, and Neutrosophic Offsets, Journal of Mathematics and Informatics, Vol. 5,
2016, 63-67, https://fs.unm.edu/SVNeutrosophicOverset-JMI.pdf
[4] Florentin Smarandache, University of New Mexico website, SuperHyperStructure &
Neutrosophic SuperHyperStructure, 2016, https://fs.unm.edu/SHS/

Critical Review. Volume XIV, 2017