NEAR GROUPS ON NEARNESS APPROXIMATION SPACES

Transcription

1 Hacettepe Journal of Mathematics and Statistics Volume 414) 2012), NEAR GROUPS ON NEARNESS APPROXIMATION SPACES Ebubekir İnan and Mehmet Ali Öztürk Received 26:08:2011 : Accepted 29:11:2011 Abstract Near set theory provides a formal basis for observation, comparison and classification of perceptual granules. In the near set approach, every perceptual granule is a set of objects that have their origin in the physical world. Objects that have, in some degree, affinities are considered perceptually near each other, i.e., objects with similar descriptions. In this paper, firstly we introduce the concept of near groups, near subgroups, near cosets, near invariant sub-groups, homomorphisms and isomorphisms of near groups in nearness approimation spaces. Then we give some properties of these near structures. Keywords: Near set, Rough set, Approimation space, Nearness approimation space, Near group AMS Classification: 03E75, 03E99, 20A05, 20E Introduction In 1982, the concept of a rough set was originally proposed by Pawlak [13] as a formal tool for modelling incompleteness and imprecision in information systems. The theory of rough sets is an etension of set theory, in which a subset of a universe is described by a pair of ordinary sets called the lower and upper approimations. A basic notion in the Pawlak rough set model is an equivalence relation. The equivalence classes are the building blocks for the construction of the lower and upper approimations. The lower approimation of a given set is the union of all the equivalence classes which are subsets of the set, and the upper approimation is the union of all the equivalence classes which have a non-empty intersection with the set. An algebraic approach to rough sets has been given by Iwinski [7]. Afterwards, Biswas and Nanda [1] introduced the notion of rough subgroups. Kuroki in [8], introduced the notion of a rough ideal in a semigroup. Kuroki and Wang [9] gave some properties of Department of Mathematics, Faculty of Arts and Sciences, Adıyaman University, Adıyaman, Turkey. E. Inan) einan@adiyaman.edu.tr M. A. Öztürk) maozturk@adiyaman.edu.tr Corresponding Author.

2 546 E. İnan, M.A. Öztürk the lower and upper approimations with respect to normal subgroups, and Davvaz [2], introduced the notion of rough subring respectively ideal) with respect to an ideal of a ring. In recent years, there has been a fast growing interest in this new emerging theory, ranging from work in pure theory such as topological and algebraic foundations [3, 4, 15, 26, 24, 25, 11, 19, 20] and [21], to diverse areas of applications [5, 6]. In 2002, near set theory was introduced by J.F. Peters as a generalization of rough set theory. In this theory, Peters depends on the features of objects to define the nearness of objects [18] and consequently, the classification of our universal set with respect to the available information of the objects. The concept of near set theory was motivated by image analysis and inspired by a study of the perception of the nearness of familiar physical objects was carried out in cooperation with Pawlak in a purely philosophical manner in a poem entitled How Near written in 2002 and published in 2007 [14]. More recent work considers a generalized approach theory in the study of the nearness of nonempty sets that resemble each other [20, 21] and a topological framework for the study of nearness and apartness of sets see, e.g., [11, 19]). Near set theory begins with the selection of probe functions that provide a basis for describing and discerning affinities between objects in distinct perceptual granules. A probe function is a real valued function representing a feature of physical objects such as images or behaviors of individual biological organisms or collections of artificial organisms such as robot societies. But in this paper, in a more general setting that includes data mining, probe functions ϕ i will be defined to allow for non-numerical values, i.e., let ϕ i : X V, where V is the value set for the range of ϕ i [22]. This more general definition of ϕ i F is also better for setting forth the algebra and logic of near sets after the manner of algebra and logic. This paper begins by introducing the basic concepts of near set theory [16]. Our aim in this article is to improve the concept of near group theory, which etends the notion of a group to include the algebraic structures of near sets. Our definition of near group is similar to the definition of rough groups [10]. Also, we introduce near subgroups, near cosets, near invariant subgroups, homomorphism and isomorphism of near groups in nearness approimation spaces, and we give some properties of these structures. 2. Preliminaries In this section we give some definitions and properties regarding near sets [16] Object Description. Table 1. Description Symbols Symbol Interpretation R Set of real numbers, O Set of perceptual objects, X X O, set of sample objects, O, sample objects, F A set of functions representing object features, B B F, Φ Φ : O R L, object description, L L is a description length, i i L, Φ) Φ) = ϕ 1),ϕ 2),ϕ 3),...,ϕ i),...,ϕ L)). 1)

3 Near Groups on Nearness Approimation Spaces 547 Objects are known by their descriptions. An object description is defined by means of a tuple of function values Φ) associated with an object X. The important thing to notice is the choice of functions ϕ i B used to describe an object of interest. Assume that B F see Table 1) is a given set of functions representing features of sample objects X O. Let ϕ i B, where ϕ i : O R. In combination, the functions representing object features provide a basis for an object description Φ : O R L, a vector containing measurements returned values) associated with each functional value ϕ i) in 1), where the description length Φ = L. Object Description : Φ) = ϕ 1),ϕ 2),ϕ 3),...,ϕ i),...,ϕ L)). The intuition underlying a description Φ) is a recording of measurements from sensors, where each sensor is modelled by a function ϕ i Nearness of Objects. Symbol Interpretation B [] B O B Table 2. Relation and Partition Symbols B= {, ) f ) = f ) f B, indiscernibility relation, [] B = { X B, elementary set class), O B= { [] B O, quotient set, ξ B Partition ξ B = O B, ϕi ϕi = ϕ i ) ϕ i), probe function difference. Sample objects X O are near each if and only if the objects have similar descriptions. Recall that each ϕ defines a description of an object see Table 1). Then let ϕi denote ϕi = ϕi ) ϕ i), where, O. The difference ϕ leads to a definition of the indiscernibility relation B introduced by Z. Pawlak [12]. See Definition 2.1) Definition. Let, O, B F. Then B= {, ) O O ϕ i B, ϕi = 0 is called the indiscernibility relation on O, where the description length i Φ Definition. Let B F be a set of functions representing features of objects, O. Objects, are called minimally near each other if there eists ϕ i B such that {ϕi, ϕi = 0. We call it the Nearness Description Principle - NDP [16] Theorem. The objects in a class [] B ξ B are near objects Near Sets. The basic idea in the near set approach to object recognition is to compare object descriptions. Sets of objects X,X are considered near each other if the sets contain objects with at least partial matching descriptions Definition. Let X,X O, B F. Set X is called near X, if there eists X, X, ϕ i B such that {ϕi Remark. If X is near X, then X is a near set relative to X and X is a near set relative to X. Notice that if we replace X by X in Definition 2.4, this leads to what is known as refleive nearness Definition. Let X O and, X. If is near, then X is called a near set relative to itself or the refleive nearness of X.

4 548 E. İnan, M.A. Öztürk 2.7. Theorem. A class in a partition ξ B is a near set Theorem. A partition ξ B is a near set Definition. Let X O, X,X X. If X, X are near sets, then X is a near set. We call it the Hierarchy of Near Sets Theorem. A set containing a near set is itself a near set Fundamental Approimation Space. This subsection presents a number of near sets resulting from the approimation of one set by another set. Approimations are carried out within the contet of a fundamental approimation space FAS = O,F, B), where O is a set of perceived objects, F is a set of probe functions representing object features, and B is an indiscernibility relation defined relative to B F. The space FAS is considered fundamental because it provided a framework for the original rough set theory [12]. It has also been observed that an approimation space is the formal counterpart of perception. Approimation starts with the partition ξ B of O defined by the relation B. Net, any set X O is approimated by considering the relation between X and the classes [] B ξ B, O. To see this, consider first the lower approimation of a set Lower Approimation of a Set. Symbol Table 3. Approimation Notation Interpretation O,F, B) Fundamental approimation space FAS), B F, B X :[] B X [] B, B-lower approimation of X, B X, B-upper approimation of X, Bnd BX :[] B X [] B Bnd BX = B X \B X = { B X and / B X. Affinities between objects of interest in the set X O and classes in the partition ξ B can be discovered by identifying those classes that have objects in common with X. Approimation of the set X begins by determining which elementary sets [] B O B are subsets of X. This discovery process leads to the construction of what is known as the B-lower approimation of X O, which is denoted by B X: B X = [] B. :[] B X Lemma. The lower approimation B X of a set X is a near set Theorem. If a set X has a non-empty lower approimation B X, then X is a near set Upper Approimation of a Set. To begin, assume that X O, where X contains perceived objects that are in some sense interesting. Also assume that B contains functions representing features of objects in O. A B-upper approimation of X is defined as follows: B X = [] B. :[] B X Theorem. The upper approimation B X and the set X are near sets.

5 Near Groups on Nearness Approimation Spaces Boundary Region. Let Bnd BX denote the boundary region of an approimation defined as follows: Bnd BX = B X \B X = { B X and / B X Theorem. A set X with an approimation boundary Bnd BX 0 is a near set Nearness Approimation Spaces. Symbol Table 4. Nearness Approimation Space Symbols Interpretation B B F, B r r B probe functions in B, Br Indiscernibility relation defined using B r, [] Br [] Br = { { O Br, equivalence class, O Br O Br = [] Br O, quotient set, ξ O,Br Partition ξ O,Br = O Br, ) r, i.e. B probe functions ϕi B taken r at a time, N r B) ν Nr N r B) X N r B) X B r N r B) = {ξ O,Br B r B, set of partitions, ν Nr : O) O) [0,1], overlap function, N r B) X = :[] Br X [] B r, lower approimation, N r B) X = :[] Br X [] B r, upper approimation, Bnd NrB)X) N r B) X N r B) X = { N r B) X / N r B) X. A nearness approimation space NAS) is a tuple NAS = O,F, Br,N r,ν Nr ) where the approimation space NAS is defined with a set of perceived objects O, a set of probe functions F representing object features, an indiscernibility relation Br defined relative to B r B F, a collection of partitions families of neighbourhoods) N rb), and a neighbourhood overlap function ν Nr. The relation Br is the usual indiscernibility relation from rough set theory restricted to a subset B r B. The subscript r denotes the cardinality of the restricted subset B r, where we consider ) B r, i.e., B functions φ i F taken r at a time to define the relation Br. This relation defines a partition of O into non-empty, pairwise disjoint subsets that are equivalence classes denoted by [] Br, where [] Br = { O Br. These classes { form a new set called the quotient set O Br, where O Br = [] Br O. In effect, each choice of probe functions B r defines a partition ξ O,Br on a set of objects O, namely, ξ O,Br = O Br. Every choice of the set B r leads to a new partition of O. Let F denoteaset of features for objects inaset X, where each φ i F maps X tosome value set V φi range of φ i). The value of φ i) is a measurement associated with a feature of an object X. The overlap function ν Nr is defined by ν Nr : O) O) [0,1], where O)isthepowersetofO. Theoverlapfunctionν Nr mapsapairofsetstoanumber in [0, 1], representing the degree of overlap between sets of objects with features defined by he probe functions B r B [23]. For each subset B r B of probe functions, define the binary relation Br = {, ) O O φ i B r, φ i) = φ i ). Since each Br is, in fact, the usual indiscernibility relation [12], for B r B and O, let [] Br denote the equivalence class containing, i.e., [] Br = { O f B r, f ) = f ). If, ) Br, then and are said to be B-indiscernible with respect to all feature

6 550 E. İnan, M.A. Öztürk probe functions in B r. Then define a collection of partitions N rb), where N r B) = {ξ O,Br B r B. Families of neighborhoods are constructed for each combination of probe functions in B using ) B r, i.e., the probe functions B taken r at a time. 3. Near groups and near subgroups 3.1. Definition. Let NAS = O,F, Br,N r,ν Nr ) be a nearness approimation space and let be a binary operation defined on O. A subset G of perceptual objects O is called a near group if the following properties are satisfied. 1),y G, y N r B) G; 2),y G, y) z = y z) property holds in N r B) G; 3) e N rb) G such that G, e = e =, e is called the near identity element of the near group G; 4) G, y G such that y = y = e, y is called the near inverse element of in G Eample. Let O = {0,a,b,c,d,e,f,g,h,ı, B = {φ 1,φ 2,φ 3 F denote a set of perceptual objects and a set of functions, respectively. Sample values of the φ 1 function φ 1 : O V 1 = {α 1,α 2,α 3, φ 2 function φ 2 : O V 2 = {α 1,α 2 and φ 3 function φ 3 : O V 3 = {α 1,α 2,α 3,α 4 are shown in Table 5. Table 5 O φ 1 φ 2 φ 3 0 α 2 α 1 α 3 a α 3 α 2 α 1 b α 2 α 1 α 3 c α 2 α 2 α 3 d α 1 α 1 α 4 e α 1 α 1 α 2 f α 3 α 2 α 2 g α 1 α 1 α 4 h α 2 α 1 α 3 ı α 3 α 2 α 1 And let be a binary operation of perceptual objects on O with the following table: Table 6 0 a b c d e f g h ı 0 0 a b c d e f g h ı a a b c d e f g h ı 0 b b c d e f g h ı 0 a c c d e f g h ı 0 a b d d e f g h ı 0 a b c e e f g h ı 0 a b c d f f g h ı 0 a b c d e g g h ı 0 a b c d e f h h ı 0 a b c d e f g ı ı 0 a b c d e f g h

7 Near Groups on Nearness Approimation Spaces 551 We can easily shown that O, ) is a group. Let G = {0,a,b,e,h,ı be a subset of the perceptual objects. Then let be a binary operation of perceptual objects on G O with the following table: Table 7 0 a b e h ı 0 0 a b e h ı a a b c f ı 0 b b c d g 0 a e e f g 0 c d h h ı 0 c f g ı ı 0 a d g h Now, Hence ξ ϕ1 ) = Thus ξ ϕ2 ) = and so ξ ϕ3 ) = [0] {ϕ1 = { O ϕ 1 ) = ϕ 10) = α 2 = {0,b,c,h = [b] {ϕ1 = [c] {ϕ 1 = [h] {ϕ 1, [a] {ϕ1 = { O ϕ 1 ) = ϕ 1a) = α 3 = {a,f,ı = [f] {ϕ1 = [ı] {ϕ 1, [d] {ϕ1 = { O ϕ 1 ) = ϕ 1d) = α 1 = {d,e,g = [e] {ϕ1 = [g] {ϕ 1. { [0] {ϕ1,[a] {ϕ 1,[d] {ϕ 1. Also, [0] {ϕ2 = { O ϕ 2 ) = ϕ 20) = α 1 = {0,b,d,e,g,h = [b] {ϕ2 = [d] {ϕ2 = [e] {ϕ 2 = [g] {ϕ 2 = [h] {ϕ 2, [a] {ϕ2 = { O ϕ 2 ) = ϕ 2a) = α 2 = {a,c,f,ı = [c] {ϕ2 = [f] {ϕ2 = [ı] {ϕ 2. { [0] {ϕ2,[a] {ϕ 2. Lastly, [0] {ϕ3 = { O ϕ 3 ) = ϕ 10) = α 3 = {0,b,c,h = [b] {ϕ3 = [c] {ϕ 3 = [h] {ϕ 3, [a] {ϕ3 = { O ϕ 3 ) = ϕ 3a) = α 1 = {a,ı = [ı] {ϕ3, [d] {ϕ3 = { O ϕ 3 ) = ϕ 3d) = α 4 = {d,g = [g] {ϕ1, [e] {ϕ3 = { O ϕ 3 ) = ϕ 3e) = α 2 = {e,f = [f]{ϕ3, { [0] {ϕ3,[a] {ϕ 3,[d] {ϕ 3,[e] {ϕ 3. Therefore, for r = 1, a classification of O is N 1B) = { ξ ϕ1 ),ξ ϕ2 ),ξ ϕ3 ).

8 552 E. İnan, M.A. Öztürk Since Then, we can write N 1B) G = :[] {ϕi G [] {ϕi = {0,b,c,h {a,f,ı {d,e,g {0,b,d,e,g,h = {0,a,b,c,d,e,f,g,h,ı = O. {a,c,f,ı {a,ı {e,f 1),y G, y N r B) G = {0,a,b,c,d,e,f,g,h,ı; 2) the property,y G, y) z = y z) holds in N r B) G; 3) 0 N r B) G such that G, 0 = 0 = 0 is called the near identity element of the near group G); 4) the properties G, y G such that y = y = e y is called a near inverse element of in G) are satisfied, the subset G of the perceptual objects O is indeed a near group Eample. Let H = {0,b,e,h be a subset of the near group G = {0,a,b,e,h,ı. Let be a binary operation of perceptual objects on H G with the following table: Table 8 0 b e h 0 0 b e h b b d g 0 e e g 0 c h h 0 c f We know from Eample 3.2, for r = 1, a classification of O is N 1B) = { ξ ϕ1 ),ξ ϕ2 ),ξ ϕ3 ). And we can write N 1B) H = :[] {ϕi H [] {ϕi = {0,b,c,h {d,e,g {0,b,d,e,g,h {e,f = {0,b,c,d,e,f,g,h O. From Theorem 3.8, since the conditions 1),y H, y N r B) H; 2) H, 1 H hold, H is a near subgroup of the near group G Eample. Let O = {0,a,b,c,d,e,f,g,h,ı, B = {φ 1,φ 2,φ 3 F denote a set of perceptual objects and set of functions, respectively. Sample values of the φ 1 function φ 1 : O V 1 = {α 1,α 2,α 3, the φ 2 function φ 2 : O V 2 = {α 1,α 2 and the φ 3 function φ 3 : O V 3 = {α 1,α 2,α 3,α 4 are as shown in Table 5. Additionally, is a binary operation of perceptual objects on O with the following table:

9 Near Groups on Nearness Approimation Spaces 553 Table 9 0 a b c d e f g h ı 0 0 a b c d e f g h ı a a b c d e f g h ı 0 b b c d e f g h ı 0 a c c d e f g h ı 0 a b d d e f g h ı 0 a b c e e f g h ı 0 a b c e f f g h ı 0 a b c d e g g h ı 0 a b c d e f h h ı 0 a b c d e f g ı ı 0 a b c e e f g h Since a d ı) a d) ı, O, ) is not a group. Let H = {0,b,e,h be a subset of the perceptual objects. Then, let be a binary operation of the perceptual objects in H O with the following table: Table 10 0 b e h 0 0 b e h b b d g 0 e e g 0 c h h 0 c f WeknowfromEample3.2, forr = 1, aclassification ofoisn 1B) = { ξ ϕ1 ),ξ ϕ2 ),ξ ϕ3 ). And, we can write N 1B) H = {0,b,c,d,e,f,g,h O. From Theorem 3.8, since the conditions 1),y H, y N rb) H; 2) H, 1 H, hold, H is a near subgroup of the near group G Proposition. Let G be a near group. 1) There is one and only one identity element in near group G. 2) G, there is only one y such that y = y = e; we denote it by 1. 3) 1) 1 =. 4) y) 1 = y Proposition. Let G be a near group. For all a,,,y,y G, 1) if a = a then =, 2) if y a = y a then = Definition. A non-empty subset H of a near group G is called its near subgroup, if it is a near group itself with respect to the operation. There is only one guaranteed trivial near subgroup of near group G, i.e. G itself. A necessary and sufficient condition for {e to be a trivial near subgroup of near group G is e G.

10 554 E. İnan, M.A. Öztürk 3.8. Theorem. A necessary and sufficient condition for a subset H of a near group G to be a near subgroup is that: 1),y H, y N r B) H; 2) H, 1 H. Proof. The necessary condition is obvious. We prove only the sufficient condition. By 1) we have,y H, y N r B) H, by 2) we have H, 1 H, so we have H, 1 N rb) H by 1) and 2). And, since association holds in N r B) G, so it holds in N r B) H. Hence the theorem is proved. An important difference between a near group and group is the following: 3.9. Theorem. Let H 1 and H 2 be two near subgroups of the near group G. A sufficient condition for the intersection of these two near subgroups of G to be a near subgroup of G is N rb) H 1 N r B) H 2 = N r B) H 1 H 2). Proof. Suppose H 1 and H 2 are two near subgroups of the near group G. It is obvious that H 1 H 2 G. Consider,y H 1 H 2. Since H 1 and H 2 are near subgroups, we have y N rb) H 1, y N rb) H 2, and 1 H 1, 1 H 2, i.e. y N rb) H 1 N r B) H 2 and 1 H 1 H 2. Assuming N r B) H 1 N r B) H 2 = N rb) H 1 H 2), we have y N rb) H 1 H 2) and 1 H 1 H 2. Thus H 1 H 2 is a near subgroup of G Definition. A near group is called a commutative near group if y = y for all,y G Eample. Eample 3.2 and Eample 3.3 are commutative near groups. 4. Near cosets Let NAS = O,F, Br,N r,ν Nr ) be a nearness approimation space, G O a near group and H a near subgroup of G. Let us define a relationship on the elements of the near group G as follows: : a b if and only if a b 1 H {e Theorem. is a compatible relation over the elements of the near group G. Proof. a G, since G is a near group, a 1 G. Since a a 1 = e, we have a a. Further, a,b G, if a b, then a b 1 H {e, i.e. a b 1 H or a b 1 {e. If a b 1 H, then, since H is a near subgroup of G, we have a b 1) 1 = b a 1 H, and thus b a. If a b 1 {e, then a b 1 = e. That means b a 1 = a b 1) 1 = e 1 = e, and thus b a. Hence, is compatible Definition. A compatible category defined by the relation is called a near right coset. The near right coset that contains the element a is denoted by H a, i.e., H a = {h a h H, a G, h a G {a. Let O,F, Br,N r,ν Nr ) be a nearness approimation space, G O a near group and H a near subgroup of G. Consider the relation on the elements of G defined as follows: : a b if and only if a 1 b H {e Theorem. is a compatible relation over the elements of the near group G.

11 Near Groups on Nearness Approimation Spaces Definition. A compatible category defined by the relation is called a near left coset. The near left coset that contains the element a is denoted by a H, i.e., a H = {a h h H, a G, a h G {a Remark. Generally speaking, the binary operation of a near group does not satisfy the commutative law, so the compatible relations and are different. As a result, the near left and right cosets are different Theorem. The near left cosets and near right cosets are equal in number. Proof. Denote bys 1, S 2 the families of near right andand left cosets, respectively. Define ϕ : S 1 S 2 such that ϕh a) = a 1 H. We prove that ϕ is a bijection. 1) If H a = H b a b), then a b 1 H. Because H is a near subgroup, we have b a 1 H, that means a 1 b 1 H, i.e. a 1 H = b 1 H. Hence, ϕ is a mapping. 2) Any element a H of S 2 is the image of H a 1, an element of S 1, hence ϕ is an onto mapping. 3) If H a H b, then a b 1 / H, i.e. a 1 H b 1 H. Hence, ϕ is a one-to-one mapping. Thus the near left cosets and near right cosets are equal in number Definition. The number of both near left cosets and near right cosets is called the inde of the subgroup H in G. 5. Near normal subgroups 5.1. Definition. A near subgroup N of a near group G is called a near normal subgroup if a N = N a for all a G Theorem. A necessary and sufficient condition for a near subgroup N of near group G to be a near normal subgroup is that a N a 1 = N for all a G. Proof. Suppose N is a near normal subgroup of G. By definition, a G we have a N = N a. Because G is a near group, we have a N) a 1 = N a) a 1, i.e. a N a 1 = N. a N a 1 = N a a 1), SupposeN isanearsubgroupofgand a G, a N a 1 = N. Then a N a 1) a = N a, i.e. a N = N a. Thus N is a near invariant subgroup of G Theorem. A necessary and sufficient condition for a near subgroup N of the near group G to be a near normal subgroup is that a n a 1 N for all a G and n N. Proof. Suppose N is anear normal subgroup ofthe near group G. Wehave a N a 1 = N for all a G. For any n N, therefore, we have a n a 1 N. Suppose N is a near subgroup of the near group G. Suppose a n a 1 N for all a G and n N. We have a n a 1 N. Because a 1 G, we further have a n a 1 N. It follows that a a 1 n a ) a 1 a n a 1, i.e. N a n a 1. Since a n a 1 N and N a n a 1, we have a n a 1 = N. Thus N is a near normal subgroup.

12 556 E. İnan, M.A. Öztürk 6. Homomorphisms of near groups Let O 1,F 1, Br1,N r1,ν Nr1 ) and O2,F 2, Br2,N r2,ν Nr2 ) be two nearness approimation spaces, and let, be binary operations over O 1 and O 2, respectively Definition. Let G 1 O 1, G 2 O 2 be near groups. If there eists a surjection ϕ : N r1 B) G 1 N r2 B) G 2 such that ϕ y) = ϕ) ϕy) for all,y N r1 B) G 1 then ϕ is called a near homomorphism and G 1, G 2 are called near homomorphic groups. Throughout this section ϕ is a near homomorphism such that ϕ : N r1 B) G 1 N r2 B) G 2, ϕ y) = ϕ) ϕy) for all,y N r1 B) G Theorem. Let G 1 and G 2 be near homomorphic groups. If satisfies the commutative law, then also satisfies it. Proof. ConsiderG 1, G 2, andϕsuchthatϕ y) = ϕ) ϕy)forall,y N r1 B) G 1. For every ϕ), ϕy) N r2 B) G 2 since ϕ is surjective, there eist,y N r1 B) G 1 such that ϕ), y ϕy). Thus ϕ y) = ϕ) ϕy), and ϕy ) = ϕy) ϕ). Now, assuming y = y, we obtain ϕ) ϕy) = ϕy) ϕ). That means that satisfies the commutative law Theorem. Let G 1 O 1, G 2 O 2 be near groups that are near homomorphic and let N r B) ϕg 1) = N r B) G 2. Then ϕg 1) is a near group. Proof. 1),y ϕg 1), consider,y G 1 such that, y y. We have ϕ y) = ϕ) ϕy) N r B) G 2 = N r B) ϕg 1), that is y N rb) ϕg 1). 2) Since e N r B) G 1, ϕe) N r B) G 2 and ϕ) ϕg 1), ϕe) ϕ) = ϕ e) = ϕ). 3) G 1 is a near group, so,y,z G 1, y z) = y) z. Hence, ϕ y z)) = ϕ) ϕy z) = ϕ) ϕy) ϕz)), ϕ y) z) = ϕ y) ϕz) = ϕ) ϕy)) ϕz), i.e., ϕ) ϕy)) ϕz) = ϕ) ϕy) ϕz)). 4) ϕg 1), consider G 1 such that. Since G 1 is a near group, 1 G 1. Hence, ϕ 1) ϕg 1) and ϕ) ϕ 1) = ϕ 1) ϕ) = ϕe). Therefore, we can put ) 1 = ϕ 1). Consequently, we can conclude that ϕg 1) is a near group Theorem. Let G 1 O 1, G 2 O 2 be near groups that are near homomorphic. Let e and e be the near identity elements of G 1 and G 2, respectively. Then ϕe) = e and ϕ a 1) = ϕa) 1, for all a N rb) G Definition. Let G 1 O 1, G 2 O 2 be near groups that are near homomorphic. Let e and e be the near identity elements of G 1 and G 2 respectively. The set { ϕ) = e, G 1 is called the near homomorphism kernel, denote by N Theorem. Let G 1 O 1, G 2 O 2 be near groups that are near homomorphic. The near homomorphism kernel N is a near invariant subgroup of G 1. Proof. Let ϕ be an onto mapping from N r1 B) G 1 to N r2 B) G 2. Then,y N we have ϕ) = e, ϕy) = e. Thus ϕ y) = ϕ) ϕy) = e e = e, i.e. y N. Moreover, N, we have ϕ) = e. Because ϕ 1) = ϕy) 1 = e 1 = e, we get 1 N. We can conclude that N is a near invariant subgroup of G 1.

13 Near Groups on Nearness Approimation Spaces Theorem. Let G 1 O 1, G 2 O 2 be near groups that are near homomorphic. Let H 1, N 1 be a near subgroup and a near normal subgroup of G 1, respectively. Then; 1) ϕh 1) is near subgroup of G 2 if ϕn r1 B) H 1) = N r2 B) ϕh 1); 2) ϕn 1) is near normal subgroup of G 2 if ϕg 1) = G 2 and ϕn r1 B) N 1) = N r2 B) ϕn 1). Proof. 1) Consider an onto mapping ϕ from N r1 B) G 1 to N r2 B) G 2 such that ϕ y) = ϕ) ϕy) for all,y N r1 B) G 1. For all ϕ), ϕy) ϕh 1), by the definition of ϕ, there eists,y H 1 such that ϕ) and ϕ) ϕy) = ϕ y) ϕn r1 B) H 1). Because ϕn r1 B) H 1) = N r1 B) ϕh 1), we have ϕ) ϕy) ϕn r1 B) H 1). Further, for all ϕ) ϕh 1), by the definition of ϕ there eists H 1 such that ϕ), y ϕy). Because H 1 is a near subgroup of G 1, we have 1 H 1. Thus ϕa) 1 = ϕ a 1) ϕh 1). We can conclude that ϕh 1) is a near subgroup of G 2. 2) By 1), it is easy to see that ϕn 1) is a near subgroup of G 2 if ϕn r1 B) N 1) = N r2 B) ϕn 1). For all ϕ) G 2, because ϕg 1) = G 2, we have ϕ) ϕg 1). Thus G 1, 1 G 1 and ϕ 1) ϕg 1) = G 2. Because for all ϕ) G 2, ϕn) ϕn 1) we have ϕ) ϕn) ϕ 1) = ϕ n 1) and N 1 is near normal subgroup of G 1, we have n 1 N 1. Hence ϕ) ϕn) ϕ 1) ϕn 1). We can conclude that ϕn 1) is a near normal subgroup of G Theorem. Let G 1 O 1, G 2 O 2 be near groups that are near homomorphic. Let H 2, N 2 be a near subgroup and a near normal subgroup of G 2, respectively. Then; 1) H 1, which is the inverse image of H 2, is a near subgroup of G 1 if ϕn r1 B) H 1) = N r2 B) H 2. 2) N 1, which is the inverse image of N 2, is a near normal subgroup of G 1 if ϕg 1) = G 2 and ϕn r1 B) N 1) = N r2 B) N 2. Proof. 1) H 1 is certainly the inverse image of H 2, and moreover we have ϕh 1) = H 2. That is, we have ϕ), ϕy) H 2 for all,y H 1. Because H 2 is a near subgroup of G 2, we have ϕ y) = ϕ) ϕy) N r2 B) H 2 = ϕn r1 B) H 1). Thus y N r1 B) H 1. We have ϕ) H 2 for all H 1. Because H 2 is a near subgroup of G 2, we have ϕ) 1 = ϕ 1) H 2. Thus 1 H 1. 2)From1), wecaneasilyshownthatn 1 isanearsubgroupofg 2 ifϕn r1 B) N 1) = N r2 B) ϕn 1). We have ϕ) ϕg 1) = G 2, ϕ) 1 = ϕ 1) ϕg 1) = G 2, ϕn) N 2 for all G 1, n N 1. Because N 2 is a near normal subgroup of G 2, we have ϕ) ϕn) ϕ 1) = ϕ n 1) N 2. Thus n 1 N 1. Hence, N 1, which is the inverse image of N 2, is a near normal subgroup of G 1 if ϕg 1) = G 2 and ϕn r1 B) N 1) = N r2 B) N Conclusion In this paper, we have studied near groups and the algebraic properties of near groups. This work is focused on near groups, near subgroups, near cosets, near invariant subgroups and homomorphism of near groups. To etend this work, one could study the properties of other algebraic structures arising from near set theory.

14 558 E. İnan, M.A. Öztürk References [1] Biswas, R. and Nanda, S. Rough groups and rough subgroups, Bull. Pol. AC. Math. 42, , [2] Davvaz, B. Rough sets in a fundamental ring, Bull. Iranian Math. Soc. 242), 49 61, [3] Davvaz, B. Roughness in rings, Inform. Sci. 164, , [4] Davvaz, B. and Mahdavipour, M. Roughness in modules, Inform. Sci ), , [5] Hassanien, A., Abraham, A., Peters, J., Schaefer, G. and Henry, C. Rough sets and near sets in medical imaging: A review, IEEE Trans Info. Tech. in Biomedicine 136), , [6] Henry, C. Near Sets: Theory and Applications Ph.D. Thesis supervisor J. F. Peters), Department of Electrical & Computer Engineering, University of Manitoba, 2010) [7] Iwinski, T. B. Algebraic approach to rough sets, Bull. Pol. AC. Math. 35, , [8] Kuroki, N. Rough ideals in semigroups, Inform. Sci. 100, , [9] Kuroki, N. and Wang, P.P. The lower and upper approimations in a fuzzy group, Inform. Sci. 90, , [10] Miao, D., Han, S., Li, D. and Sun, L. Rough Group, Rough Subgroup and their Properties Springer-Verlag, Heidelberg, 2005), [11] Naimpally, S. and Peters, J. Topology with applications. Topological spaces via near and far, World Scientific, Singapore, 2012 to appear). [12] Pawlak, Z. Classification of Objects by means of Attributes, Institute for Computer Science, Polish Academy of Sciences, Report 429, [13] Pawlak, Z. Rough sets, Int. J. Comput. Inform. Sci. 115), , [14] Pawlak, Z., Peters, J.F. and Blisko, J. How near, Systemy Wspomagania Decyzji I 57109), [15] Pawlak, Z. Rough Sets-Theoretical Aspects of : Reasoning about Data Kluwer Academic Puplishers, Boston, London, Dordrecht, 1991). [16] Peters, J.F. Near sets. General theory about nearness of objects, Applied Mathematical Sciences ), , [17] Peters, J. F. Near sets, special theory about nearness of objects, Fund. Inform ), , [18] Peters, J. F. Classification of perceptual objects by means of features, Int. J. Info. Technol. Intell. Comput. 32), 1 35, [19] Peters, J.F. Sufficiently near sets of neighbourhoods, in: J. Yao, S. Ramanna, G. Wang, Z. Suraj eds.), Rough Sets and Knowledge Technology, LNCS 6954, Springer, Berlin, 17 24, [20] Peters, J. F. and Naimpally, S. Approach spaces for near filters, Gen. Math. Notes 21), , [21] Peters, J.F. and Tiwari, S. Approach merotopies and near filters, Gen. Math. Notes 31), 1 15, [22] Polkowski, L. Rough Sets Mathematical Foundations, Springer-Verlag, Heidelberg, 2002). [23] Skowron, A. and Stepaniuk, J. Tolerance approimation spaces, Fund. Inform ), , [24] Wolski, M. Perception and classification. A note on near sets and rough sets, Fundamenta Informaticae 101, , [25] Wolski, M. Gauges, pregauges and completions: Some theoretical aspects of near and rough set approaches to data, in: J. Yao, S. Ramanna, G. Wang, Z. Suraj eds.), Rough Sets ad Knowledge Technology, LNCS 6954, Springer, Berlin, , [26] Yao, Y. Y. On generalizing Pawlak approimation operators, Lecture Notes in Artificial Intelligence 1424, , 1994.