INTRODUCTION

Rough set theory, proposed by Zdzislaw Pawlak (1982), models uncertainty

by equivalence relations. The primary notion is the partitioning of the domain into
equivalence classes. Objects belonging to the same equivalence cannot be
distinguished. Hence Rough set theory is a theory of multiple memberships.

A graph is a convenient way of representing information involving

relationships between objects. The objects are represented by vertices and relations
by edges. When there is vagueness in the description of the objects or in their
relationships or in both, it is natural that we need to design a “Rough graph model”.

An attribute set was first assigned to the edges of a graph. Given a subset R
of the attribute set, there is associated an equivalence relation, based in which R-
rough is defined. If the vertices of a graph are connected through edges denoting
different possible relationship.

In this paper the core is taken up the central structure of a rough graph &
weighted rough graph. The core of graph G is a path of minimum distance. Minieka
& Patel have investigated the problem of finding a tree core with a specified length.
Eliezer. A, Albacea has presented a parallel algorithm for finding the core of the tree
with weights only on edges.

In a particular application, villages may be considered vertices and roads as

edges. In such cases the population of a village will then be the rough weight of the
vertex, the bus route, which goes adjacent to the villages may be treated as central
structure now the people of several villages have to reach the bus route traversing
the lowest possible distance.

1
 BASIC DEFINITION
DEFINITION 1.1

ROUGH SET

A Rough Set is a formal approximation of a crisp-set in terms of a pair of sets

which the lower and upper approximation of the original set.

Let 𝑈 denote the set of objects called Universe and let 𝑅 be an equivalence
relation on 𝑈. The pairs 𝐴 = (𝑈, 𝑅) is called an Approximation Space. For 𝑢, 𝑣 ∈
𝑈, 𝑢 and belong to the same equivalence class and we say that they are
indistinguishable in 𝐴. The relation 𝑅 is called an Indiscernibility Relation.

Let [𝑋]𝑅 denote an equivalence class of 𝑅 containing element 𝑥, then lower

& upper approximation for a subset 𝑋⊆𝑈 in 𝐴 denoted by
𝐴(𝑋) & 𝐴(𝑋) respectively,

Where 𝐴(𝑋) = {𝑥 ∈ 𝑈/[𝑋]𝑅 ⊂ 𝑋}

𝐴(𝑋) = {𝑥 ∈ 𝑈/[𝑋]𝑅 ∩ 𝑋 ≠ ∅} Thus, if an object 𝑥 ∈ 𝐴(𝑋) then “𝑥 surely

belongs to 𝑋 in 𝐴” If 𝑥 ∈ 𝐴(𝑋) then “𝑥 possibly belong to 𝑋 in 𝐴”.

DEFINITION:1.2

MEMBERSHIP VALUE
|𝐴(𝑋)|
The membership value of 𝑋 is 𝜇(𝑋) =
|𝐴(𝑋)|

#([𝑋]𝑅 ∩𝑋)
The membership value of each element of 𝑋 in 𝐴 is 𝜇𝑋 (𝑥) =
#[𝑋]𝑅

2
RESULTS:1.1

1. If 𝐴(𝑋) = 𝐴(𝑋) then 𝑋 is definable on attribute set 𝐴.

2. If 𝐴(𝑋) ≠ 𝐴(𝑋) then 𝑋 is undefinable (roughly definable) on attribute set 𝐴.

i. If 𝐴(𝑋) ≠ ∅ & 𝐴(𝑋) ≠ 𝑈 then 𝑋 is undefinable (R- roughly definable) on

attribute set 𝐴.
ii. If 𝐴(𝑋) ≠ ∅ & 𝐴(𝑋) == 𝑈 then 𝑋 is internally undefinable (𝑅𝑖 ) on attribute
set 𝐴.
iii. If 𝐴(𝑋) = ∅ & 𝐴(𝑋) ≠ 𝑈 then X is externally undefinable (𝑅𝑒 ) on attribute
set 𝐴.
iv. If 𝐴(𝑋) = ∅ & 𝐴(𝑋) = 𝑈 then 𝑋 is totally undefinable (𝑅𝑡 ) on attribute set 𝐴.

DEFINITION:1.3

CRISP GRAPH

A Rough graph is a pair 𝑅(𝐺) = (𝑅(𝐺), 𝑅(𝐺) where 𝑅(𝐺): (𝑉, 𝐸) is a crisp
graph and 𝑅(𝐺): (𝑉, 𝐸) is a fuzzy graph and 𝑅(𝐺) ⊆ 𝑅(𝐺)

A weighted rough graph is of the form 𝐺: (𝑅, 𝑅, 𝑉, 𝜎, 𝜇) Where 𝜇 is a fuzzy

subset of a nonempty set 𝐸 & 𝜎 is a fuzzy relation on 𝜇.

Define 𝜇1 : 𝑅 → (0,1), 𝜇2 : 𝑅 → 1 &

𝜎(𝑢) ≥ max {𝜇(𝑢, 𝑢𝑖 )/𝑢 𝑎𝑛𝑑 𝑢𝑖 𝑎𝑟𝑒 𝑎𝑑𝑗𝑎𝑐𝑒𝑛𝑡, 𝑤ℎ𝑒𝑟𝑒 1 ≤ 𝑖 ≤ 𝑛}.

3
DEFINITION:1.4

COMPLEMENT OF A WEIGHTED ROUGH GRAPH

The complement of a weighted rough graph G: (𝑅, 𝑅, V, 𝜎, 𝜇) is Gˈ:(𝑅 ˈ, 𝑅

ˈ Vˈ, 𝜎 ˈ, 𝜇ˈ) where,

(i) Vˈ = V
(ii) 𝜇ˈ =1- 𝜇
(iii) 𝜎 ˈ(𝑢) ≥ {max {𝜇ˈ (𝑢,𝑢𝑖 ) / 𝑢 and 𝑢𝑖 are adjacent where 𝑢𝑖 ∈ V}.

DEFINITION:1.5

DEGREE

Let G be a weighted rough graph. The Degree of an edge 𝑒 = 𝑢𝑣 is

𝑑𝑅(𝐺) (𝑢𝑣) =𝜎(𝑢)+𝜎(𝑣). Minimum degree of G is 𝛿(𝐺)= ∧{𝑑(𝑢𝑣)/𝑢𝑣 ∈ 𝐸}.
Maximum degree of G is △ (𝐺)= ∨{𝑑(𝑢𝑣)/𝑢𝑣 ∈ 𝐸}. Let G be a weighted rough
graph. The degree of an vertex ‘𝑣’ is defined by 𝑑(𝑅(𝐺)) (𝑣) = ∑ 𝜇 (𝑢, 𝑣)
𝑢≠𝑣

DEFINITION: 1.6

K- REGULAR WEIGHTED EDGE ROUGH GRAPH.

Let 𝑅(G) :(𝜎, 𝜇) be a rough graph. If 𝑑𝑅(𝐺) (𝑒) = 𝑘 for all e ∈ 𝐸, then 𝑅G) is
said to be an regular weighted edge rough graph of degree k or a k- regular
weighted edge rough graph.

4
REMARK:

(𝑅(G)) is a k-edge regular weighted rough graph if and only if 𝛿=△=k.

DEFINITION:1.7

WEIGHTED ROUGH GRAPH

A weighted rough graph is strong if 𝜎(𝑢) = max {𝜇 (𝑢,𝑢𝑖 ) / 𝑢 and 𝑢𝑖 are

adjacent, where 1 ≤ 𝑖 ≤ n}.

DEFINITION:1.8

COMPLETE

A weighted rough graph is complete if 𝜎(𝑢) = max {𝜇 (𝑢, 𝑢𝑖 ) / 𝑢 and 𝑢𝑖 are

adjacent, ⩝ 𝑖 = 1,2 … 𝑛} 𝑢, 𝑢1 𝑢2 ,…, 𝑢𝑛 ∈ σ* and every Adjacent Edges 𝑢𝑢1 ,
𝑢𝑢2 ,…,𝑢𝑢𝑛 ∈ μ* where G:( σ*, μ*) is a crisp graph.

DEFINITION:1.9

LENGTH & STRENGTH

A path P of length n is a sequence of distinct nodes 𝑢0 𝑢1 , … , 𝑢𝑛 s.t 𝜇 (𝑢𝑖−1 ,

𝑢𝑖 )>0, 𝑖=1,2…n & the degree of membership of a weakest edge is defined as its
strength. The strength of connectedness between two nodes u & v is defined as the
maximum of the strength of all path between u & v.

5
DEFINITION:1.10

CONNECTED ROUGH GRAPH

Let 𝑅(𝐺) = (𝑅(𝐺), 𝑅(𝐺) be a connected rough graph. If there is at most one
strongest path between any two nodes of R(G), then R(G) is a rough graph tree.

DEFINITION: 1.11

UNIVERSE GRAPH

Let 𝐺 = (𝑉, 𝐸) be a universe graph with 𝑉 = {𝑣1 𝑣2 … 𝑣𝑛 } and 𝐸 =

{𝑒1 𝑒2 , … , 𝑒𝑚 }. Let 𝑅 be an equalence relation defined on 𝑉. Then the elements in 𝑉
can be divided into different equivalence classes [𝑣]𝑅 .

DEFINITION: 1.12
R-DEFINABLE GRAPH

Let T (W, X) be a sub-graph of G (V, E) where W ⊆ V, X ⊆ E, graph T is

called R-definable graph or R-exact graph if W is the union of some [𝑣]𝑅.
Otherwise, graph T is called R-undefinable graph or R-rough graph.

DEFINITION: 1.13
Let T (W2, X) and S (W2, Y) are subgraphs of G (V, E) where W1 ⊆ V, W2
⊆ V, X ⊆ E, Y ⊆ E, T = (R*(T), R*(T)) and S = (R*(S), R*(S)) be its rough graphs.
S is said to be surely subgraph of Tif R*(S) ⊆ R*(T). Also, S is said to be possibly
subgraph of T if R*(S) ⊆ R*(T) if S is both surely subgraph and possibly subgraph
of T, then S is a rough subgraph of T.
6
DEFINITION:1.14
MULTIPLE OR PARALLEL EDGES
A set of two or more edges of a rough graph T is said to be multiple or
parallel edges if they have the same end vertices. An edge for which two ends are
the same is called a loop at the common vertex.

A rough graph T = (R*(T), R*(T)) is said to be surely simple if R*(T) contains

no loops and parallel edges.

A rough graph T = (R*(T), R*(T)) is said to be possibly simple if R*(T)

contains no loops and parallel edges. A rough graph T = (R*(T), R*(T)) is said to
be simple if it is both surely and possibly simple graphs.

DEFINITION: 1.15
+(𝑣) = max{∆𝔎 (𝑣𝑖 ) ∕ 𝑣𝑖 ∈ 𝑉(𝐺)}
−(𝑣) = min{∆𝔎 (𝑣𝑖 ) ∕ 𝑣𝑖 ∈ 𝑉(𝐺)}
If + (v) = - (v), then the graph is said to be regular rough graph.

DEFINITION: 1.16
COMPLETE ROUGH GRAPH
If every vertex of rough graph is adjacent with all other vertices, then it is said
to be complete rough graph. The complete rough graph with 𝑛 vertices is denoted
by ℑn.

7
DEFINITION: 1.17
ROUGH WALK
An alternating sequence of vertices and edges in a rough graph with
𝜔(𝑣i) ≥ 0.5, is said to be rough walk.

DEFINITION: 1.18
Distinct edges in rough walk, is said to be rough trial and distinct vertices in
a rough walk, is said to be rough path. It is denoted by 𝔓𝑛 .

DEFINITION: 1.19
A rough cycle is defined as the closed rough walk 𝑣1 , 𝑣2 , … , 𝑣𝑛 = 𝑣1 , where
n≥ 3 and 𝑣1 , 𝑣2 , 𝑣𝑛−1 are distinct. It is denoted by ℭn.

8
 CORE IN ROUGH GRAPH

DEFINITION:2.1

R-ROUGH GRAPH

Let U= (V, E) be the universal graph. Where V= {𝑣1 , 𝑣2 , 𝑣3 , … , 𝑣𝑛 } is the set

of vertices & E= {𝑒1 , 𝑒2 , 𝑒3 …𝑒𝑛 } is the set edges on U where the edge 𝑒𝐾 is endued
with vertex attribute (𝑣𝑖 , 𝑣𝑗 ). Let R = {𝑟1 𝑟2 𝑟3 … 𝑟|𝑅| } be the attribute set on U. For any
attribute set R⊆R on E, the elements of E can be divided into different equivalence
classes [𝑒]𝑅 .

For any graph T= (W, X), where W⊆V and X⊆ E, If X is the sum of
equivalence classes, then the graph is called R-definable graph or R-exact graph. If
not, the graph is called R-un definable graph or R-rough graph. For R-rough
graph, two exact graphs,

𝑅(𝑇) = (𝑊, 𝑅(𝑋)) & 𝑅(𝑇) = (𝑊, 𝑅(𝑋)) can be used to define it approximately.

Where 𝑅(𝑋) = {𝑒 ∈ 𝐸 ⁄[𝑒]R ⊂ 𝑋}

𝑅(𝑋) = {𝑒 ∈ 𝐸 ⁄[𝑒]R∩ 𝑋 ≠ ∅}. The graph 𝑅(𝑇) & 𝑅(𝑇) are called R-lower
and R-upper approximation of graph T. The pair of Graph (𝑅(𝑇), 𝑅(𝑇)) is called
R-rough graph.

RESULTS: 2.1

1. The graph T =(𝑅(T), 𝑅(T)) is exact if and only if (𝑅(T)= 𝑅(T)).

9
 2. The graph T =(𝑅(T), 𝑅(T)) is called a rough graph if and only if (𝑅(T)≠ 𝑅(T).

3. The graph T =(𝑅(T), 𝑅(T)) is a classical graph if and only if all the edges of

T belong to the same equivalence class with respect to R.

EXAMPLE:1

Let us consider the graph G= (W, X) where W⊆V and X⊆ E

W= {𝑣1, 𝑣2 , 𝑣3 , 𝑣4 , 𝑣5 , 𝑣6 , 𝑣7 , 𝑣8 , 𝑣9 }and

E= {𝑒1 , 𝑒2 , 𝑒3 , 𝑒4 , 𝑒5 , 𝑒6 , 𝑒7 , 𝑒8 , 𝑒9 , 𝑒10 }.

𝑒1 𝑒2

𝑒3
𝑒4 𝑒5 𝑒6

𝑒7
𝑒8
𝑒9
𝑒10

Figure 2.1

Let X= {𝑒1 , 𝑒3 , 𝑒4 , 𝑒6 , 𝑒10 }. Let R be the equivalence relation, here R=edge

coloring.

The equivalence classes are {{𝑔, 𝑔, 𝑔}, {𝑟, 𝑟}, {𝑦, 𝑦, 𝑦}, {𝑏, 𝑏}}

E= {{e1 , e6 , e10 }, {e2 , e3 }, {e4 , e5 , e9 }, {e7 , e8 }}

10
R-Lower & R-Upper approximation of graph G are 𝑅(𝐺)& 𝑅(𝐺) can be used to
define it approximately.

where 𝑅(X)= {e ∈ E/[𝑒]𝑅 ⊂ X}

= {𝑒1 , 𝑒6 , 𝑒10 }

𝑅(X)= {e ∈ E/[𝑒]𝑅 ⋂ X≠ 𝜙}

= {𝑒1 , 𝑒2 , 𝑒3 , 𝑒4 , 𝑒5 , 𝑒6 , 𝑒9 , 𝑒10 }
#([𝑒1 ]𝑅 ∩𝑋)
The membership values are 𝜇𝑋 (𝑒1 ) = = 1,
#([𝑒1 ]𝑅

𝜇𝑋 (𝑒6 ) = 𝜇𝑋 (𝑒10 ) = 1

1
𝜇𝑋 (𝑒2 ) = 𝜇𝑋 (𝑒3 ) =
2
1
𝜇𝑋 (𝑒4 ) = 𝜇𝑋 (𝑒5 ) = 𝜇𝑋 (𝑒9 ) =
3

RESULTS:2.2

1. 𝑅(𝑋) ⊂ 𝑋 ⊆ 𝑅(𝑋)

2. 𝑅(∅) ⊂ ∅ ⊆ 𝑅(∅)

3. 𝑅(𝑈) ⊂ 𝑈 ⊆ 𝑅(𝑈)

4. 𝑅(𝑋 ∩ 𝑌) = 𝑅(𝑋) ∩ 𝑅(𝑌)

5. 𝑅(𝑋 ∪ 𝑌) = 𝑅(𝑋) ∪ 𝑅(𝑌)

6. 𝑋 ⊆ 𝑌 ⇒ 𝑅(𝑋) ⊆ 𝑅(𝑌)& 𝑅(𝑋) ⊆ 𝑅(𝑌)

11
 7. 𝑅(𝑋 ∪ 𝑌) ⊇ 𝑅(𝑋) ∪ 𝑅(𝑌)

8. 𝑅(𝑋 ∩ 𝑌) ⊆ 𝑅(𝑋) ∩ 𝑅(𝑌)

9. 𝑅 (𝑅(𝑋)) = 𝑅 (𝑅(𝑋)) = 𝑅(𝑋)

10.𝑅 (𝑅(𝑋)) = 𝑅 (𝑅(𝑋)) = 𝑅(𝑋)

2.1 ROUGH -SPANNING SUB GRAPH

THEOREM:2.1

Let G be a complete graph with 𝑛 ≥ 2 vertices then

(i) Number of definable spanning sub graph under minimum edge coloring is

2𝑛 , 𝑖𝑓 𝑛 𝑖𝑠 𝑜𝑑𝑑
𝑛(𝑑(𝐾𝑛 )) = { 𝑛+1 }
2 , 𝑖𝑓 𝑛𝑖𝑠 𝑒𝑣𝑒𝑛

(ii) Number of undefinable (rough) spanning sub graph under minimum edge
coloring Is

𝑛(𝑛−1)
2 2, − 2𝑛 𝑖𝑓 𝑛 𝑖𝑠 𝑜𝑑𝑑
𝑛(𝑅 (𝐾𝑛 )) = { 𝑛(𝑛−1) }
2 2 − 2𝑛−1 , 𝑖𝑓 𝑛 𝑖𝑠 𝑒𝑣𝑒𝑛

12
 THEOREM:2.2

Let G be a cycle with 𝑛 ≥ 3 vertices then

(i) Number of definable spanning sub graph under minimum edge coloring is

23 , 𝑖𝑓 𝑛 𝑖𝑠 𝑜𝑑𝑑
𝑛(𝑑(𝐾𝑛 )) = { 2 }
2 , 𝑖𝑓 𝑛𝑖𝑠 𝑒𝑣𝑒𝑛

(ii) Number of undefinable (rough) spanning sub graph under minimum edge

coloring is

2𝑛 − 23 , 𝑖𝑓 𝑛 𝑖𝑠 𝑜𝑑𝑑
𝑛(𝑅 (𝐶𝑛 )) = { 𝑛 }
2 − 22 , 𝑖𝑓 𝑛 𝑖𝑠 𝑒𝑣𝑒𝑛

THEOREM:2.3

Let 𝐺(𝑉, 𝐸) be any graph. If the number of edges is equal to the number the
equivalence classes under attribute R then all the spanning sub graph of the given
graph is definable under the attribute R.

EXAMPLE: 2

1. In 𝑲𝟑 : All the edge sub graph of 𝐾3 are definable under minimum edge

coloring 𝑛(𝑑(𝐾4 )) = 23

2. In 𝑲𝟒 : n (d (𝐾4 )) = 23, n (R (𝐾4 )) = 12, n (𝑅𝑖 (𝐾4 )) = 18, n (𝑅𝑒 (𝑲𝟒 )) = 18

n (𝑅𝑡 (𝑲𝟒 )) =8.

13
 3. In 𝑲𝟓 : n (d (𝐾4 )) = 25, n (R (𝐾4 )) =620, n (𝑅𝑖 (𝐾4 )) =170, n (𝑅𝑒 (𝐾4 )) =172

n (𝑅𝑡 (𝐾4 )) =30.

4. In C4: n (d (C4)) = 22, n (R (C4)) =0, n (𝑅𝑖 (C4)) =4, n (𝑅e (C4)) =4

n (𝑅t (C4)) =4.

5. In C5: n (d (C5)) =23, n (R (C5)) =8, n (𝑅𝑖 (C5)) =8 n (𝑅e (C5)) =8

n (𝑅t (C5)) =0.

2.2 DISTANCE IN ROUGH GRAPH

DEFINITION:2.2

ECCENTRICITY

Eccentricity of a vertex v, denoted by 𝑒R(v) is the distance from v to a vertex

farthest from it

𝑒𝑅 (𝑣) = 𝑚𝑎𝑥{𝑑𝑅 (𝑢, 𝑣): 𝑢 ∈ 𝑉},

𝑑𝑅 (𝑢, 𝑣) =∧𝑝 {𝑙𝑝 × 𝑠𝑝 }.

𝑤ℎ𝑒𝑟𝑒 𝑝 𝑖𝑠 𝑎 (𝑢, 𝑣)𝑝𝑎𝑡ℎ,

𝑙𝑝 − 𝑙𝑒𝑛𝑔𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑎𝑡ℎ

𝑠𝑝 − 𝑠𝑡𝑟𝑒𝑛𝑔𝑡ℎ 𝑜𝑓 𝑎 𝑝𝑎𝑡ℎ

14
 DIAMETER

Diameter is the maximum eccentricity & radius is the minimum eccentricity.

Each vertex in V at which the eccentricity function is minimized is called a center.
A vertex v is called peripheral vertex if 𝑒𝑅 (𝑣) = 𝑑𝑖𝑎𝑚 (𝑅(𝐺)). Average distance is
1
defined by A(R(G)) = ∑ 𝑑𝑅 (𝑢, 𝑣). Each vertex in V at which the distance function
𝑛 𝑐2

is minimized is called a median of R(G). The path of minimum distance is called a

core (or) path median of R(G).

THEOREM:2.4

In a rough graph 𝑅(𝐺), 𝑑𝑅 : 𝑉 × 𝑉 → [0,1] is a metric on 𝑉, ∀ 𝑢, 𝑣, 𝑤 ∈ 𝑉

(i) 𝑑R (𝑢, 𝑣) ≥ 0 ⩝ (𝑢, 𝑣) ∈ V

(ii) 𝑑R (𝑢, 𝑣) = 0 𝑖𝑓𝑓 u=v
(iii) 𝑑R (𝑢, 𝑣) = 𝑑R (𝑣, 𝑢)
(iv) 𝑑R (𝑢, 𝑤) ≤ 𝑑R (𝑢, 𝑣) + 𝑑R (𝑣, 𝑤).

THEOREM:2.5

Let 𝑅(𝐺)= (𝑅(𝐺), 𝑅(𝐺)) be a complete rough graph then diam (𝐾𝑛 ) ≤ 1&
rad (𝐾𝑛 ) < 1.

15
THEOREM:2.6

Let 𝑅(𝐺)= (𝑅(𝐺), 𝑅(𝐺)) be a complete bipartite rough graph then

diam (𝐾𝑚,𝑛 ) ≤ 2 & rad (𝐾𝑚,𝑛 )<2.

THEOREM:2.7

Let 𝑅(𝐺)= (𝑅(𝐺), 𝑅(𝐺)) be a path on n vertices then

diam (𝑃𝑛 ) = ∑𝑛𝑖=1 𝜇(𝑢1 , 𝑢𝑖 )

rad (𝑃𝑛 ) ≤ ∑𝑛𝑖=1 𝜇(𝑢1 , 𝑢𝑖 )

THEOREM:2.8

Let 𝑅(𝐺)= (𝑅(𝐺), 𝑅(𝐺)) be a complete rough graph & rad(R(G) =1 then
R(G) is a self-centered graph.

2.3 FIND THE CORE OF THE GIVEN ROUGH GRAPH TREE

Finding the core of the given tree consists of the following four steps.

1. Find the end point for the given tree; call it e1.

2. Re-root the tree at e1.

3. Find the end point of the new tree; call it e2.

4. The path from e1 to e2 forms the core of the tree.

16
 These steps are explained in detail with the help of an example. Finding the
endpoint of a core includes the following four steps.

(i) For each vertex 𝑣 ∈ 𝑉, compute the size (number of nodes) of the sub tree

rooted at 𝑣, size (𝑣).

(ii) For each vertex v ∈ V, compute reduction (𝑟, 𝑣).

(iii) Compute max r =max {reduction (𝑟, 𝑣); 𝑣 ∈ 𝑉.

(iv) The end point is the vertex v with reduction (𝑟, 𝑣)= max 𝑟

1
0.5 1

2
0.7 3
0.6 0.3
0.7
4
5 6
7
1
0.5
8 Figure 2.2
9

17
Step 1: find the end point of the given tree

(i) Compute size (v)

𝑉 1 2 3 4 5 6 7 8 9

Size 9 4 4 2 1 2 1 1 1

(ii) Compute reduction (𝑟, 𝑣) where 𝑟 = 1.

Reduction:

(𝑟, 𝑢2 ) = .5 × 4 = 2,

(𝑟, 𝑢3 ) = 1 × 4 = 4,

(𝑟, 𝑢4 ) = 2 × .7 + 4 × .5 = 3.4,

(𝑟, 𝑢5 ) = 1 × .6 + .5 × 4 = 2.6

(𝑟, 𝑢6 ) = 2 × .3 + 4 × 1 = 4.6

(𝑟, 𝑢7 ) = 1 × .2 + 4 × 1 = 4.7

(𝑟, 𝑢8 ) = 1 × 1 + 2 × .7 + .5 × 4 = 4.4

(𝑟, 𝑢9 ) = 1 × .5 + 2 × .3 + 1 × 4 = 5.1

(iii) Since the highest reduction is 5.1, the corresponding vertex 9 is the end
point, 𝑒1 = 9

18
 Step 2: Re – root the tree 𝑒1

9
0.5

6
0.3

3
1
0.7
1
0.5 7

2
0.7
0.6

4
5
1
Figure 2.3
8

Step 3: find the end point of the re – rooted tree.

(i) Compute size (v)

𝑉 1 2 3 4 5 6 7 8 9

Size 5 4 7 2 1 8 1 1 9

19
 (ii) Compute reduction (𝑟, 𝑣) where 𝑟 = 9.

Reduction:

(𝑟, 𝑢6 ) = 4,

(𝑟, 𝑢3 ) = 4 + 2.1 = 6.1,

(𝑟, 𝑢7 ) = 6.1 + 0.7 = 6.8,

(𝑟, 𝑢1 ) = 6.1 + 5 = 11.1,

(𝑟, 𝑢2 ) = 13.1,

(𝑟, 𝑢4 ) = 13.7

(𝑟, 𝑢5 ) = 14.1

(𝑟, 𝑢8 ) = 15.5

(iii) Since the highest reduction is 15.5, the corresponding vertex 8 is the end
point 𝑒2 = 8

Step 4: the path 𝑒1 to 𝑒2 forms the core of the tree

20
 1
0.5 1

2
0.7 3
0.6 0.3
0.7
4
5 6
7
1 0.5

8
9

Figure 2.4

21
 CORE IN WEIGHTED ROUGH GRAPH

3.1 DISTANCE IN WEIGHTED ROUGH GRAPH

DEFINITION 3.1

Eccentricity of a vertex v, denoted by 𝒆𝒘𝑹 (𝒗) is the distance from v to a

vertex farthest from it 𝒆𝒘𝑹 (𝒗) = 𝒎𝒂𝒙{𝒅𝒘𝑹 (𝒖, 𝒗): 𝒖 ∈ 𝑽}, 𝒅𝒘𝑹 (𝒖, 𝒗) =

𝒅𝑹 (𝒖, 𝒗) × 𝝈(𝒗) ∀ 𝒖, 𝒗 ∈ 𝑽 𝑤ℎ𝑒𝑟𝑒 𝑑𝑅 (𝑢, 𝑣) is distance in rough graph diameter is

the maximum eccentricity & radius is the minimum eccentricity .Each vertex in V

at which the eccentricity function is minimized is called a center of R(G).A vertex

v is called peripheral vertex if e(v)=diam(R(G).

1
Average distance is defined by 𝐴(𝑅(𝐺)) = ∑ 𝑑𝑤𝑅 (𝑢, 𝑣)
𝑛 𝑐2

Each vertex in V at which the distance function is minimized is called a

median. The path of minimum distance is called a w-core (or) path median.

THEOREM 3.1:

For any weighted rough graph G, the radius and diameter satisfy

rad(G)≤diam(G)≤ 2rad(G).

22
THEOREM 3.2

Let G be a weighted rough graph, center of graph need not be same as the
center of its underlying graph.

THEOREM 3.3

For any two adjacent vertices 𝑢, 𝑣 in a weighted rough graph |𝒆𝒘𝑹 (𝒖) −
𝒆𝒘𝑹 (𝒗)| ≤ 1

THEOREM 3.4

For any two vertices 𝑢, 𝑣 in a weighted rough graph |𝒆𝒘𝑹 (𝒖) − 𝒆𝒘𝑹 (𝒗)| ≤
𝒅𝒘𝑹 (𝒖, 𝒗)

REMARK 3.1

In ordinary trees, the center is 𝐾1 (or) 𝐾2 , but in weighted rough graph trees,
the center need not be 𝐾1 (or) 𝐾2 .

23
 a

c e

d f
g
Figure 3.1

Center of a tree in crisp graph

0.4

0.2 b 0.1

c e
0.1
0.1
1
d f
g
Figure 3.2

Center of a tree in rough graph

24
 0.4

0.4

0.4
0.2 0.1

0.2 1
0.1 0.1
1
0.1 0.1
1
Figure 3.3

Center of a tree in weighted rough graph

REMARK 3.2

The center of a rough graph and a weighted rough graph are need not be same.

In fig 3.2, the central vertices of a rough graph are a, b, c, d &f. But in fig 3.3,
the cental vertices of a weighted rough are b &f.

THEOREM 3.5

A rough graph which is self-centered need not be a self-centered weighted

rough graph.

25
3.2 FIND THE CORE OF THE GIVEN WEIGHTED
ROUGH GRAPH TREE
Finding the core of the given tree consists of the following four steps.

1. Find the end point for the given tree; call it 𝑒1 .

2. Re-root the tree at 𝑒1 .

3. Find the end point of the new tree; call it 𝑒2 .

4. The path from 𝑒1 to 𝑒2 forms the w - core of the tree.

These steps are explained in detail with the help of an example. Finding
the end point of a core includes the following four steps.

i. For each vertex v ∈ V, compute the sub tree weighted distance

at v, 𝑠𝑡𝑤𝑟𝑑 (v).
ii. For each vertex v ∈ V, compute reduction (𝑟, 𝑣).
iii. Compute max 𝑟 =max {reduction (𝑟, 𝑣); v ∈ V.
iv. The end point is the vertex v with reduction (𝑟, 𝑣)= max 𝑟.

0.5 1(1)
1

2(.7) 3(1)
0.7
0.7
0.3
4(1) 0.6
6(.8) 7(.7)

5(.6)
0.5
1
Figure 3.4
9(.5)
8(1)

26
Step 1: find the end point of the given tree

(i) compute size (v)

𝑉 1 2 3 4 5 6 7 8 9

Size - 2.76 1.04 1 0.6 0.25 0.7 1 0.5

(ii) Compute reduction (𝑟, 𝑣) where 𝑟 = 1.

Reduction:

(𝑟, 𝑢2 ) = 1.38, (𝑟, 𝑢3 ) = 1.04, (𝑟, 𝑢4 ) = 2.08,

(𝑟, 𝑢5 ) = 1.74, (𝑟, 𝑢6 ) = 1.79, (𝑟, 𝑢7 ) = 1.53, (𝑟, 𝑢8 ) = 3.08, (𝑟, 𝑢9 ) = 1.94

(iii) Since the highest reduction is 3.08, the corresponding vertex 8 is the end
point 𝑒1 = 8

Step 2: Re – root the tree 𝑒1

27
 8(1)
1

4(1)
0.7

2(.7)
0.5 0.5

1(1) 5(.6)
1

3(1)
0.3
0.7
6(.5) 0.5
9(.5)
7(.7)

Figure 3.5

Step 3: find the end point of the re – rooted tree.

(i) Compute sub tree weighted distance (𝑠𝑡𝑤𝑟𝑑)

𝑉 1 2 3 4 5 6 7 8 9

Size 2.79 0.36 1.04 7.8 0.6 0.25 0.7 − 0.5

28
 (ii) Compute reduction (𝑟, 𝑣) where 𝑟 = 8.

Reduction: (𝑟, 7) = 10.537 since the highest reduction is 10.537 the

corresponding vertex 7 is the end point, 𝑒2 = 7

Step 4: the path 𝑒1 to 𝑒2 forms the core of the tree

1(1)
0.5 1

2(.7)
0.7 3(1)
0.6 0.3
0.7
4(1)
5(.6) 6(.5)
7(.7)
1 0.5

8(1)
9(.5)

Figure 3.6

29
 VERTEX ROUGH GRAPH

DEFINITION:4.1

R-vertex rough graph is defined in terms of two exact graphs R*(T) =

(R*(W), R∗(X)) and R*(T) = (R*(W), R*(X)), where

R*(W), = {v ∈ V: [v]R ⊆ W}

R*(W), = {v ∈ V: [v]R ∩ W = φ}

R∗(X) = {(vi , vj) ∈ X: vi , vj ∈ [v]R for some v ∈ s R*(W)}

R∗(X) = (vi,vj) ∈ E: vi ∈ [vi]R & [vi]R ∩ X=φ and v j ∈ [vj]R & [vj]R ∩ X =
φ} .

The graphs R*(T) and R*(T) are called R-lower approximate graph of T and R-
upper approximate graph of T. The pair of graphs (R*(T), R*(T)) is called R-vertex
rough graph.

EXAMPLE:1

Consider G (V, E) V = {v1, v2, v3,

v4, v5} V/R = {{v3, v4, v5},

{v1, v2,}}

Consider T = (W, X) be a subgraph of G (V, E) (Fig. 1). we get the lower and
upper approximations of vertex set and edge set as

R*(X) = {e1, e2, e3, e4, e5, e6}, R*(T) = (R*(W), R*(X)), R*(T) = (R*(W), R*(X))

30
 𝑣5 𝑒4 𝑣4 𝑣5 𝑒4 𝑣4

𝑒5
𝑒5 𝑒3
𝑒6 𝑒3

𝑒1
𝑣1
𝑣2 𝑒2 𝑣3 𝑒2
𝑣2 𝑣3

𝐺(𝑉, 𝐸) 𝑇(𝑊, 𝑋)

Fig 4.1 graph 𝑮(𝑽, 𝑬) and its sub graph 𝑻(𝑾, 𝑿)

PROPOSITION:4.1.1

Lower and upper approximations of a graph have the

following properties: For all T, T1, T2 ⊆ G,

1. R* (T) ⊆ T ⊆ R*(T).

2. R*(Kc) = R*(Kc) = Kc, R∗(G) = R*(G) = G where K is the Complete

graph.

3. R*(T1 ∩ T2) = R∗(T1) ∩ R∗(T2).

4. R*(T1 𝖴 T2) = R*(T1) 𝖴 R*(T2).

5. R∗ (T1 𝖴 T2) ⊇ R∗(T1) 𝖴 R∗(T2).

6. R*(T1 ∩ T2) ⊆ R*(T1) ∩ R*(T2).

7. T1 ⊆ T2 ⇒ R∗(T1) ⊆ R∗(T2) & R*(T1) ⊆ R*(T2).

31
Example :2

Consider G (V, E)

V = {v1, v2, v3, v4, v5, v6, v7},

V/R = {{v7} {v4, v5}, {v1, v2}, {v3, v6}}

Consider T = (W, X) be a subgraph (Fig. 3) of G (V, E) Then, we get the lower

approximation of vertex set and edge set as (Fig. 4)

R∗(W) = {v3, v4, v5, v6} R(X)∗ = {e4, e5}

R∗(T ) = (R∗(W), R∗(X))

Here R∗(T ) is disconnected. but T = (W, X) is connected. Also T is a tree but

R∗(T ) is not a tree. It is a forrest.

𝑣5 𝑒4 𝑣4

𝑅∗ (𝑇) 𝑒3

𝑣3

32
 𝑣5 𝑒4 𝑣4

𝑅∗ (𝑇)

𝑒5 𝑒3
𝑒6

𝑒1
𝑣2 𝑒2 𝑣3
𝑣1

𝑒1 𝑣2
𝑣1

𝑒2
𝑒3 𝑣4 𝑒4
𝑣3 𝑣5

𝑒5 𝑒6

𝑣6 𝑣7

𝑅∗ (𝑇)

33
 APPLICATIONS

• 2D MATRIX GAMES:

In 2d matrix games can be used to find the optimal path for maximum sum
along starting to end points and many variations of it can be found online.

• SPANNING TREES:

Weighted graphs are used to find the minimum spanning tree from graph
which depicts the minimal cost to traverse all nodes in the graph.

• CONSTRAINTS GRAPHS:

Graphs are often used to represent constraints among items. Used in

scheduling, product design, asset allocation, circuit design, and artificial
intelligence.

• DEPENDENCY GRAPHS:

Directed weighted graphs can be used to represent dependencies or

precedence order among items. Priority will be assigned to provide a flow in
which we will solve the problem or traverse the graph from highest priority to
lowest priority. Such graphs are often used in large projects in laying out what

34
components rely on other components and are used to minimize the total time or
cost to completion while abiding by the dependencies.

• COMPILERS:

Weighted graphs are used extensively in compilers. They can be used for
type inference, for so-called data flow analysis, and many other purposes such as
query optimization in database languages.

• ARTIFICIAL INTELLIGENCE:

Weighted graphs are used in artificial intelligence for decision-making

processes, such as in game trees for determining the best move in a game.

• IMAGE PROCESSING:

Weighted graphs are used in image processing for segmentation, where the
weight of the edges represents the similarity between two pixels.

• NATURAL LANGUAGE PROCESSING:

Weighted graphs are used in natural language processing for text

classification, where the weight of the edges represents the similarity between
two words.

35
• TRANSPORTATION NETWORKS:

Using weighted graphs, we can figure things out like the path that takes
the least time, or the path with the least overall distance. This is a simplification
of how weighted graphs can be used for more complex things like a GPS system.
Graphs are used to study traffic patterns, traffic light timings and much more by
many big tech companies such as OLA, UBER, RAPIDO, etc. Graph networks
are used by many map programs such as Google Maps, Bing Maps, etc.

• DOCUMENT LINK GRAPHS:

Link weighted graphs are used to analyze relevance of web pages, the best
sources of information, and good link sites by taking the count of the number of
views as weights in the graph.

• EPIDEMIOLOGY:

Weighted graphs can be used to find the maximum distance transmission

from an infectious to a healthy person.

• GRAPHS IN QUANTUM FIELD THEORY:

Vertices represent states of a quantum system, and the edges represent

transitions between them. The graphs can be used to analyze path integrals and

36
summing these up generates a quantum amplitude. Research to find maximum
frequency along a path can be done using weighted graphs.

• SOCIAL NETWORK GRAPHS:

We can find which all users are connected in a network both directly
(direct connection) and indirectly (indirect connection). But now weighted
graphs are also used in social media for many purposes, for example, in recent
times Instagram is using features like close friends which is not the same as all
friends these features are being implemented using weighted graphs.

• NETWORK PACKET TRAFFIC GRAPHS:

Network packet traffic graphs are used for analyzing network security,
studying the spread of worms, and tracking criminal or non-criminal activity.

37
 CONCLUSION

When comparing fig 3.5 & fig 3.6 the core of a rough tree is different from
w-core of the weighted rough tree. Hence the vertex weights have a direct impact on
the decision of selecting a facility location.

We make a further study to the combination of rough set and graph theory.
Graph theory possesses a sophisticated mathematical structure. And it has been
widely used in the real world. So, we hope our work in this paper could contribute
to the theoretical development and applications of rough set.

38
 BIBLIOGRAPHY

1. Abd-El Aziz E. Radwan, Mohammed Shokry Nayle, Ahmed Ibrahem Nasir,

New rough set properties on graph theory, Int. J. Contemp. Math. Sciences,

7 (2012), no. 25, 1217 – 1232.

2. E.A. Albacea, Parallel algorithm for finding a core of a tree network,

Information Processing Center, 51 (1994), 223-226.

http://dx.doi.org/10.1016/0020-0190(94)00102-2.

3. Anjan Mukherjee, S. Bhattacharya Halder, Fuzzy Sets and Fuzzy Topology,

2015.

4. Frank Harary, New Directions in the theory of Graphs, Academic Press,

1971. [8]. Zdzislaw Pawlak, Rough sets, Int. Journal of Computer and

Information Sciences, 11 (1982), 341-356.

http://dx.doi.org/10.1007/bf01001956 Received: March 24, 2016; Published:

June 1, 201

5. Fred Bukley, Frank Harary, Distance in Graphs, Addision –Wesly, Red

Wood city, CA, 1990. 1 3(1) 7 1 6(.5) 8 5 1 4 1 2 1 8(1) 9(.5) 4(1) 7(.7))

5(.6) 2(.7) 1(1) 1)1.

39
6. E. Minieka and N.H, Patel, On finding the core of a tree with a specific

length, Journal of Algorithm, 4 (1983), 345-352.

http://dx.doi.org/10.1016/0196-6774(83)90015-9.

7. A Rosenfield, Fuzzy Graphs, Chapter in In Fuzzy Sets and Their

Applications to Cognitive and Decision Processes, L.A. Zadeh, K. S. Fu, M.

Shimura, (Eds) Academic Press, New York, 1975, 77-95.

http://dx.doi.org/10.1016/b978-0-12-775260-0.50008-6.

40