|
Academic Open Internet Journal |
Volume 11, 2004 |
Some properties of mappings
in the model VH
Gokhan Çuvalcıoğlu
University of Mersin,
Faculity of Arts and Science
Department of Mathematics
e-mail: gcuvalcioglu@mersin.edu.tr
23 August 2003
Abstract: In this paper we study the basic decomposition theorem for mappings in VH that is the model of intuitionistic set theory which is an extension of fuzzy set theory
Keywords: intuitionistic set, intuitionistic relation, intuitionistic mappings.
1. Introduction:
The notion of a fuzzy set was first introduced in 1965 by Zadeh [5]. The theory has found wideranging applications in such diverse fields as automata, control theory, decision theory, and social behavior pattern studies, to name just a few. At the same time, a fuzzy relation between X and Y as a fuzzy subset of X´Y was also proposet by Zadeh [5].
In [5], the fuzzy sets was
introduced as an extension of crisp (usual) sets, the usual two-valued sets in
ordinary set theory, by enlarging the truth value set to the real unit interval
[0,1]. Fuzzy sets are characterized by mappings called membership
functions into [0,1]. The basic fuzzy
set operations (union, intersection, complement, Cartesian product, etc.) are
defined by some equations or inequalities between the membership functions.
Intuitionistic fuzzy sets
constitute a generalisation of the notion of a fuzzy set and were introduced by
Atanassov [1].
The model VH
called Heyting model for intuitionistic set theory, where H is a complete
Heyting algebra and is considered as the set of truth values in the model. The
fuzzy sets and fuzzy relation was interpretted by shimoda [3]. It was indicated
that VH is a model for intuitionistic set theory by Titani [4].
In this paper, H is
Heyting valued set with the classical operations, Ù, Ú, Ù, Ú, ®, Ø, 0, 1, £ .
In intuitionistic logic,
the strictness axioms: xÎy®ExÙEy and x=y®ExÙEy. The theory of intuitionistic logic
with existence predicate is developed in Scott [2].
Let V be the class of all
usual sets and On be the class of ordinals. The model VH
is constructed as follows.
Definition 1: For every aÎOn,
= Æ
= ![]()
= {u = <ïuï,Eu> : ïuï: Du®H, Du Í
, EuÎH, ïuï(x) £ Eu Ù Ex, for all xÎDu}
then the H-valued model is
VH = ![]()
If
uÎVH then u is called a set in VH.
We identifiy ïuï with u.
Let
j(a1, a2, a3, ... ,an) be a
formula and j be a sentence. A formula of VH is a
formula of intuitionistic set theory with constants from elements of VH.
Definition2: Let u, vÎVH, j and y be sentences of VH and j(a) be a formula of VH. Then
i-
úçEuúç = Eu
ii-
úçuÎvúç =
)
iii-
úçu = vúç =
) Ù
) Ù Eu Ù Ev
iv-
úçj Ù yúç = úçjúçÙúçyúç
v-
úçj Ú yúç = úçjúçÚúçyúç
vi-
úçj ® yúç = úçjúç®úçyúç
vii-
úçØjúç = Øúçjúç
viii-
úç"j(x) úç =
úçj(x) úç)
ix-
úç$j(x) úç =
Ùúçj(x) úç)
We abbreviate "x(xÎu®j(x)) to ("xÎu)j(x) and $x(xÎuÙj(x)) to ($xÎu)j(x) resp.
Lemma 1: Let j(a) be a formula of VH and uÎ VH.
i-
úç"xj(x) úç =
úçj(x) úç) =
úçj(x) úç)
ii-
úç$xj(x) úç =
úçj(x) úç) =
úçj(x) úç)
Thus for u, vÎ VH,
úçu Í vúç =
úçxÎvúç) =
úçxÎvúç)
Let u, vÎVH. If úçuÍvúç=1 then we say u is a subset of v in VH and write uôv. If uôv and v ô u then ve say u and are similar and denote it as u:v.
Lemma 2: Let u, v Î VH
i-
=
Ù
Ù Eu Ù Ev
ii- u
ôv iff
£
for all x Î VH
iii- u
~ v iff
=
for all x Î VH
Definition 3: For u, v Î VH the cartesian product of u and v in VH
is defined in the following way,
D(u
v) = {
: x Î Du, y Î Dv}, E(u
v) = Eu Ù Ev
and (u
v) :
→
u(x) Ù v(y)
We will write
instead of
.
From the above definition we write,
=
Ù
Proposition 1: All axioms of intuitionistic set theory are
valid in VH.
From the above proposition, it is
show that VH is a model of intuitionistic set theory.
Definition 4: Let u, v Î VH be sets. The
intersection and union of u and v are defined as following.
i-
D(u
v) = Du
Dv, E(u
v) = Eu Ù Ev
and u
v : x →
Ù![]()
ii-
D(u
v) = Du
Dv,
E(u
v) = Eu Ú Ev and u
v : x →
Ú![]()
Lemma 3: For all x, u, v Î VH , the followings
hold.
i-
=
Ù![]()
ii-
=
Ú![]()
Lemma 4: For all u, v, w Î VH ,
i-
u
v ô u, u
v ô v and if w ô u and w ô v then w ô u
v
ii-
u ô u
v, u ô u
v and
if u ô w and v ô w then u
v ô w
2. Relations in the model VH
Definition 5: The relation in VH is a
subset of a cartesian product of two sets in VH.
For
R, u, v Î VH , R is a relation from u to v in
VH iff R
u´v.
If u = v then R is called the relation on u. We often write xRy instead of
ÎR.
Definition 6: Let u Î VH be a set. The
identity relation Iu on u defined by
D(Iu) = {
: xÎDu},
E(Iu) = Eu and Iu:
→u(x)
Iu is also called the equality
relation on u.
Lemma 5: Let u Î VH be a set.
i-
=
Ù![]()
ii-
=
, for
all zÎVH
iii-
Iu
ô u ´ u, that is Iu is a
relation on u.
Definition 7: For R,S ÎVH, the composition SoR
in VH is defined by
D(SoR) = {
: x,z Î VbH}, E(RoS) = ERÙES and SoR:
→ ![]()
where b Î On such that DR
DS
VbH
Lemma 6: Let R, S, u, v, w ÎVH be sets. Then the
following holds.
i-
=
, for
all x,y Î VH.
ii-
=
, for
all t Î VH.
iii-
If R ô u ´ v and Sôv ´
w then SoR ô u ´ w.
iv-
(ToS)oR
= To(SoR).
v-
If R ô u ´ v then RoIu ~ R~IvoR.
Definition 8: For R Î VH, the inverse relation
R-1 of R in VH is definite by
D(R-1) = {
: x,y Î VH}, E(R-1) = ER
and R-1 :
→
.
where b is an
ordinal which satisfies DR
VbH.
Lemma 7: Let R, u, v, then the following holds.
i-
.
ii-
=
, for
all w Î VH.
iii-
If R ô u ´
v then R-1 ô v ´
u.
iv-
(u
´ v)-1 ~ v ´ u.
Proposition 2: Let R,S, u,v Î VH be sets then the following
conditions are hold.
i-
(
)-1 ô R, if
R ô u´v then (
)-1 ~ R.
ii-
~
Ç
.
iii-
~
o
.
iv-
If R ô S then
ô
.
The most of the properties of relations on VH
are defined same way as in V.
Definition 9: u Î VH and R be a relation on u in
VH.
i-
R is
reflexive iff úç ("xÎu)(xRx) úç=1.
ii-
R is
symmetric iff úç ("x"y)(xRy ® yRx) úç=1.
iii-
R is
transitive iff úç ("x"y"z)(xRyÙyRz ® xRz) úç=1.
iv-
R is
connected iff úç ("x"y)(xRy Ú yRx) úç=1.
The relation on a set is equivalence relation in VH
iff it is reflexive, symmetric and transitive. For aÎu, the set
is defined by
D(
) = {y:yÎ
}, E
=Ea, and
:x®R(a,x)
Thus, we define the set U/R with,
D(u/R)={
:aÎu}, E(u/R) = EuÙER, and u/R:
®![]()
It is called the quotient set of u by R.
Lemma 8: Let R be a
relation on uÎ VH
i-
R is
reflexive iff Iu ô R.
ii-
R is
symmetric iff
ô R iff R ~
.
iii-
R is
connected iff u´v ô RÈ
iff u´u : RÈ
.
Definition 10: Let R be a relation from u to v in VH
i-
R is
total if úç ("xÎu)($yÎv)(xRy) úç=1.
ii-
R is
surjective if úç ("yÎv)($xÎu)(xRy) úç=1.
iii-
R is
injective if úç ("x,yÎu)("zÎv)(xRz Ù yRz ® x = y) úç=1.
iv-
R is
univalent if úç ("xÎu)("y,zÎv)(xRy Ù xRz ® y = z) úç=1.
For each uÎVH, the identity relation Iu
on u is defined by:
D(Iu)={<xx>: xÎDu}, E(Iu)=Eu and Iu :
<xx>®u(x)
Proposition 3: Let R be a relation from u to v in VH.
i-
R is
total iff Iu ô
oR.
ii-
R is
surjective iff Iv ô Ro
.
iii-
R is
injective iff
oRôIu.
iv-
R is
univalent iff Ro
ôIv.
Proof: Let R be a relation from u to v in
VH.
i-
Since
= úç ("xÎu)x(
oR)x úç=úç ("xÎu)($yÎv) (xRy) úç, R is total iff
=1 iff Iu
ô
oR.
ii-
In VH,
oRÍIu iff ("x, yÎu)(x(
oR)y®x=y) iff ("x, yÎu)($zÎv)(xRzÙyRz)®x=y) iff R is injective.
iii and iv similar to i and iii respectively.
It is obvious that the inverse of a total (resp. surjective) relation is surjective (resp. total) and the inverse of an injective (resp. univalent) relation is univalent (resp. injective). A composition of total (resp. surjective, injectivei univalent) relations is total (resp. surjective, injective, univalent)
3. Mappings in model VH
Definition 11: Let u, v, fÎ VH, f is called a mapping from
u to v in VH if f is total and univalent relation from u to v in VH.
We write f:u®v in VH if f is a mapping from u tov in VH. Hence f:u®v in VH iff úçf:u®vúç=1, that is,
úçf Í u´vúç=úç ("x$y)(<xy>Îfúç = úç ("x"y"z)(<xy>ÎfÙ<xz>Îf)®(y=z) úç=1.
We often write f(x) = y
instead of <xy> Î f or xfy. By proposition1.2.(1)(4), a
relation f from u to v is a mapping iff Iu ô
and
ô Iv .
A composition of two
mappings is also a mappings, that is, if f is a mapping from u to v and g is a
mapping from v to w, then the composition gof is a mapping from u to w in VH.
Identity relations are always mappings.
For a mapping f from u to
v in VH, the composition Rf =
is always an equivalence relation. Then we have úçxRfyúç = úçf(x)=f(y) úç for all x, yÎV, where, f(x) = f(y) is an abbreviation
for $z(xfz Ù yfz).
An injection (resp.
Surjection) is an injective (resp. Surjective) mapping, and a bijection is an
injective and surjective mapping. By proposition 3 we have
Lemma 9: Let f be a relation from u to v in VH
i-
f is an
injection iff Iu :
and
ô Iv.
ii-
f is a
surjection iff Iu ô
and
: Iv.
iii-
f is bijection
iff Iu :
and
: Iv.
Hence for a mapping f:u®v in VH, f is injective iff Iu
:
and is surjective iff
: Iv.
Theorem 1: Let u,vÎ VH be sets, f:u®v is a function in VH then there is a decomposition of f such
that
f ~ kohog
where k:f(u)®v is an injection, h:
®f(u) is a bijection and g:u®
is a surjection.
Proof: Let úç<a
>Îgúç=úç<ab>ÎIuúç=úça=búç then
úç
Í
úç=úç ("
,
Î
)(<![]()
>Î
®
=
úç
=
úç ("![]()
$d)(<d
>Îg Ù <d
>Îg ®
=
úç
=
úç ("![]()
$d)( úçd=búç Ù úçd=cúç ®
=
úç=1
therefore
ô
.
úçg Í u´
úç= úç ("a$
)(<a
> Îgúç= úç ("a$![]()
)(<a
>ÎgÙ<a
>Îg®
=
úç=1.
From
the same way, if we define the function h:
®f(u) with
úç<
y>Îhúç= ![]()
we obtained h is a bijection.
If we choose k is a embedding mapping then it is
clear that k is an injection.
We proved only
f ô kohog,
úçfÍ kohogúç=
(f(x,y)®(x,y)Î kohog)
=
(f(x,y)®($zÎf(u), k(z,y)Ù(x,z)Î hog)
=
(f(x,y)® ($zÎf(u),
Î
k(z,y)Ù(x,
)Îg(
,z)Î h)
=
(f(x,y)® ($zÎf(u),
Î
,(z=y)Ù(x=a)Ù
((m=a)Ù(m,z)Îf))
³
(f(x,y)® (($z,
),(z=y)Ù(x=a)Ù(a,z)Îf)
=
(f(x,y)®f(x,y)) = 1
therefore f ô kohog.
The function g is called
natural mapping which is defined above theorem and write natRf
instead of g.
Theorem 2: Let u,vÎVH be sets. R be an equivalence
relation on u, and f:u®v is a function in VH. If RôRf then there exist a function
:
®v in VH such that
o natRf :f.
Proof: We define úç<
,y>Î
úç=
úç<by>Îfúç. From this
úç<x,y>Î
onatRúç= úç$
, <x,
>ÎnatRÙ<
y>Î
úç
=
(úç
ÎA/RúçÙúç<x,
>ÎnatRúçÙúç<
y>Î
úç)
£
(úç
ÎA/RúçÙúçx=aúçÙúç<ay>Îfúç)
£
(úçaÎAúçÙúçx=aúçÙúç<ay>Îfúç)
£ úç<xy>Îfúç.
Therefore, f ô
onatR.
The converse is clear from the definition of
úçfÍ
onatRúç=
(<xy>Îf®<xy>Î
onatR).
[1]
Atanassov, Intuitionistic fuzzy sets, Phiysica-Verlag,
Heidelberg, New York, 1999.
[2] Scott, Identity
and existence in intuitionistic logic. In: MP Fourman, et. al., Application of
Sheaves., 753 (1979), Lecture Notes in Mathematics, 660-696
[3] Shimoda,
M., A natural interpretation of fuzzy sets and fuzzy relations, (128)
(2002), Fuzzy Sets and Systems, 135-147.
[4] Titani,
S., Completeness of global intuitionistic set theory, 62 (2) (1997) J.
Symbolic Logic, 506-528.
[5] Zadeh, L.A., Fuzzy sets, Inform and
Control 8 (3) (1965) 338-353.
Technical College -
Bourgas,
All rights reserved, ©
March, 2000