Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

An Approach for Solving Multi-Objective Linear

Fractional Programming Problem with fully Rough
Interval Coefficients

Mohamed Solomon 1, Hegazy Zaher 2, Naglaa Ragaa 3

1
Ph.D. Researcher, Department of Operations Research, Faculty of Graduate Studies for
Statistical Research, Cairo University, Egypt.
2
Professor Doctor in Mathematical Statistics, Faculty of Graduate Studies for Statistical
Research, Cairo University, Egypt.
3
Professor Doctor in Operations Research, Faculty of Graduate Studies for Statistical
Research, Cairo University, Egypt.

Abstract
In this paper, a multi-objective linear fractional programming (MOLFP) problem is
considered where all of its coefficients in the objective function and constraints are rough
intervals (RIs). At first to solve this problem, we will construct two MOLFP problems with
interval coefficients. One of these problems is an MOLFP where all of its coefficients are
upper approximations of RIs and the other is an MOLFP where all of its coefficients are lower
approximations of RIs. Second, the MOLFP problems are transformed into a single objective
linear programming (LP) problem using a proposal given by Nuran Guzel. Finally the single
objective LP problem is solved by a regular simplex method which yields an efficient solution
of the original MOLFP problem. A numerical example is given to demonstrate the results.
Keywords: rough set theory, multi-objective, interval coefficients, rough interval
coefficients, multi-objective linear fractional programming, linear fractional programming.

1. Introduction
In several applications of nonlinear programming a function is to be maximized or minimized
which involves one or several ratios of functions. Such optimization problems are commonly
called fractional programs, abbreviating the term 'fractional functionals program' initially
suggested by Charnes and Cooper [2] in their classic paper in 1962. Rarely the term
'hyperbolic program' is used as well. To improve the terminology one may think of the term
'ratio program.' However, such a change may not come easy after well over one thousand
publications have appeared in this area of nonlinear programming.

Linear Fractional Programming (LFP) problem is a mathematical programming problem

where the objective function is the ratio of two linear functions subject to the constraints with
linear equalities or inequalities. The Hungarian mathematicians, Martos and Whinston, [4]
developed linear fractional programming problem in the 1960s. LFP problem is applied when
the constraints and objective functions are deterministic in nature.

Volume 23, Issue 7, July - 2021 Page -94

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

Fractional programming involves the optimization of one or several ratios of functions subject
to some linear restrictions. In literature, various methods can be observed to solve different
models of linear fractional programming LFP problem. Among the solution methods, the
transformation technique developed by Charnes and Cooper [2], the simplex based algorithm
proposed by Swarup [13] is widely accepted. The simplex method for solving LFP problem
described by Bajaliov [9] is similar to Swarup’s method.

The Rough set theory approach has major importance in the areas of machine learning,
knowledge acquisition, decision analysis, and knowledge discovery from a database [11]. It
has been successfully applied in many real-world problems, counting decision algorithms
[25], pharmacology [12], and civil engineering [17] and among others. at the current time,
some papers have been established on rough programming [3]. Newly, a new kind of rough
programming was suggested by Youness [8] and Osman et al. [15], where they defined two
solutions, ideas as a surely optimal solution and possibly optimal solution.
In 2006, Robolledo [16] suggested RIs and then the rough intervals used to deal with
partially unknown or ill-defined parameters and variables. RIs are presented to adjust the
rough set principles to model continuous variables. RSTs were used only to handle discrete
objects, initially, and could not represent continuous values. RI is a specific case of rough
sets. It achieves all the rough sets’ properties and basic concepts, including the upper and
lower approximation definitions [16]. More details of RIs are stated in the next section.
In the modern age, some new approaches have also been reported to solve MOLFP and
FMOLFP problems. Farhana Akond [10] developed a method for solving fuzzy multi-
objective linear fractional programming FMOLFP problem. At first the FMOLFP problem is
converted into (crisp) multi-objective linear fractional programming MOLFP problem using
the graded mean integration, representation (GMIR) method proposed by Chen and Hsieh.
That is, all the fuzzy parameters of FMOLFP problem are converted into crisp values. Then
the MOLFP problem is transformed into a single objective linear programming LP problem
using a proposal given by Nuran Guzel [18]. Finally the single objective LP problem is solved
by a regular simplex method which yields an efficient solution of the original FMOLFP
problem. Jain [22] proposed a method using Gauss elimination technique to derive a
numerical solution of multi-objective linear programming (MOLP) problem. Then Jain [23]
extended his work for MOLFP problem. Porchelvi et al. [20] presented procedures for solving
both MOLFP problem and FMOLFP problem using the complementary development method
of Dheyab [1], where the fractional linear programming is transformed into a linear
programming problem. Guzel and Sivri [19] presented a method for finding an efficient
solution of MOLFP problem using goal programming. Later Guzel [18] proposed a simplex
type algorithm for finding an efficient solution of MOLFP problem based on a theorem
studied in a work by Dinkelbach [24], where he converted the main problem into a single
objective LP problem.

Ammar and muamer [6] proposed algorithm for solving fuzzy rough linear fractional
programming problem, where all variables and coefficients are fuzzy rough number. After
that they used the decomposition to the fuzzy linear fractional programming problem for
obtaining an optimal fuzzy rough solution, based on the variable transformation method.
Further, the proposed approach can be extended for solving FRLFP problem where all
coefficients are trapezoidal fuzzy numbers. Then Ammar and muamer [7] extended their work
by presented a new approach for solving multi-objective linear fractional programming with
fuzzy rough coefficients (MOFRLFP) problem by two methods (𝛼 − 𝑐𝑢𝑡, ranking function).
Later Mohamed S. Osman et al. [14] proposed approach to solve multi-level multi-objective

Volume 23, Issue 7, July - 2021 Page -95

 Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

fractional programming problem where some or all of its coefficients in the objective function
are rough intervals. in the first phase of the solution approach and to avoid the complexity of
the problem, two FP problems with interval coefficients will be constructed. At the second
phase, a membership function was constructed to develop a fuzzy goal programming model
for obtaining the satisfactory solution of the multi-level multi-objective fractional
programming problem. In 2021, E. Fathy [5] presented a suitable solution procedure to solve
the fully rough multi-objective multi-level linear fractional programming (FRMMFP)
problem. First, an extension of interval method is presented to deal with the roughness of the
stated problem. Then, an iterative technique is proposed for linearization of fractional
objectives. Finally, a modification of fuzzy approach is provided in the environment of the
fully rough to solve the linear model.
The motivation of our discussion in this paper is to improve a method to determine the
optimal solution of an MOLFP problem with rough interval coefficients.
The rest of the paper is prepared as follows. In Section 2, some basic knowledge of RI`s are
presented. In Section 3, an MOLFP problem is discussed. In Section 4, an MOLFP problem
with rough interval coefficients is discussed. Section 5, proposed a solution method for an
MOLFP problem with rough interval coefficients. In section 6, numerical example for
illustrating the solution of proposed method. Finally, concluding remarks are given in Section
7.

2. Rough intervals
in this section, Some definitions and properties of rough intervals are given. [16]

Definition: The qualitative value 𝐴 is called a rough interval when one can assign two closed
intervals 𝐴∗ and 𝐴∗ on R to it where 𝐴∗ ⊆ 𝐴∗ . Moreover,
(a) If 𝑥 ∈ 𝐴∗ then 𝐴 surely takes 𝑥 (denoted by 𝑥 ∈ 𝐴).
(b) If 𝑥 ∈ 𝐴∗ then 𝐴 possibly takes 𝑥.
(c) If 𝑥 ∉ 𝐴∗ then 𝐴 surely does not take 𝑥 (denoted by 𝑥 ∉ 𝐴).
𝐴∗ and 𝐴∗ are called the lower approximation interval (LAI) and the upper approximation
interval (UAI) of 𝐴, respectively. Additional, 𝐴 is denoted by 𝐴 = (𝐴∗ , 𝐴∗ ).
Note that the intervals 𝐴∗ and 𝐴∗ are not the complement of each other.
The arithmetic operations on RIs are based on interval arithmetic [16]. We will state some of
these arithmetic operations as follows [12]:

Let 𝐴 = ( [𝑎𝑙 , 𝑎𝑢 ], [𝑎𝑙 , 𝑎𝑢 ]) and 𝐵 = ( [𝑏 𝑙 , 𝑏 𝑢 ], [𝑏 𝑙 , 𝑏 𝑢 ]) be two rough intervals. Then, we

have:

[Addition] 𝐴 + 𝐵 = ([𝑎𝑙 + 𝑏 𝑙 , 𝑎𝑢 + 𝑏 𝑢 ], [𝑎𝑙 + 𝑏 𝑙 , 𝑎𝑢 + 𝑏 𝑢 ]).

[Subtraction] 𝐴 − 𝐵 = ([𝑎𝑙 − 𝑏 𝑢 , 𝑎𝑢 − 𝑏 𝑙 ], [𝑎𝑙 − 𝑏 𝑢 , 𝑎𝑢 − 𝑏 𝑙 ]).

[Negative] −𝐴 = ([−𝑎𝑢 , −𝑎𝑙 ], [−𝑎𝑢 , − 𝑎𝑙 ]).

[Intersection] 𝐴 ∩ 𝐵 = ([max{𝑎𝑙 , 𝑏 𝑙 } , min{𝑎𝑢 , 𝑏 𝑢 }], [max {𝑎𝑙 , 𝑏 𝑙 } , min{𝑎𝑢 , 𝑏 𝑢 }].

Volume 23, Issue 7, July - 2021 Page -96

 Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

[Union] 𝐴 ∪ 𝐵 = ([min{𝑎𝑙 , 𝑏 𝑙 } , max{𝑎𝑢 , 𝑏 𝑢 }], [min {𝑎𝑙 , 𝑏 𝑙 } , max{𝑎𝑢 , 𝑏 𝑢 }].

3. Multi-Objective Linear Fractional Programming Problem

An MOLFP problem is defined as follows

(𝑴𝑶𝑳𝑭𝑷) 𝑴𝒂𝒙𝒊𝒎𝒊𝒛𝒆 {𝒁(𝒙) = (𝒛𝟏 (𝒙), 𝒛𝟐 (𝒙), … , 𝒛𝒌 (𝒙))}

𝒔. 𝒕 𝑨𝒙 ≤ 𝒃 (1)
𝒙 ≥ 𝟎.
Where:
𝑺 = {𝒙𝝐ℝ𝒏 | 𝑨𝒙 ≤ 𝒃, 𝒙 ≥ 𝟎, 𝒃𝝐ℝ𝒎 } , is the Feasible Set in Decision Space.
𝑨 𝒊𝒔 𝒂𝒏 𝒎 × 𝒏 𝒎𝒂𝒕𝒓𝒊𝒙, 𝒙 ∈ ℝ𝒏 𝒂𝒏𝒅 𝒃 ∈ ℝ𝒎 ; (𝒃 ≥ 𝟎), 𝒌 ≥ 𝟐.
𝒄𝑻𝒊 𝒙 + 𝜶𝒊 𝑵𝒊 (𝒙)
𝒛𝒊 (𝒙) = 𝑻 = ; 𝒄𝑻𝒊 , 𝒅𝑻𝒊 𝝐ℝ𝒏 ; 𝜶𝒊 , 𝜷𝒊 𝝐ℝ; 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑖 = 1,2, … , 𝑘
𝒅𝒊 𝒙 + 𝜷𝒊 𝑫𝒊 (𝒙)
𝒂𝒏𝒅 𝑫𝒊 (𝒙) = 𝒅𝑻𝒊 𝒙 + 𝜷𝒊 > 𝟎, 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑖 = 1,2, … , 𝑘, 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 𝜖 𝑆.

̅𝝐 𝑺 is an efficient solution of the problem (MOLFP) if and only if there is no

A solution 𝒙
𝒙 𝝐 𝑺 𝑠𝑢𝑐ℎ 𝑡ℎ𝑎𝑡 𝒛𝒊 (𝒙) ≥ 𝒛𝒊 (𝒙
̅) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑖 = 1,2, … , 𝑘 𝑎𝑛𝑑 𝒛𝒊 (𝒙) > 𝒛𝒊 (𝒙
̅) 𝑓𝑜𝑟 𝑎𝑡 𝑙𝑒𝑎𝑠𝑡 𝑜𝑛𝑒 𝑖.
Note that, for vectors 𝑎𝑛 ; 𝒙 ≥ 𝒚 𝑖𝑚𝑝𝑙𝑖𝑒𝑠 𝒙𝒊 ≥ 𝒚𝒊 𝑓𝑜𝑟 𝑒𝑎𝑐ℎ 𝒊 , 𝒙 ≥ 𝒚 𝑖𝑚𝑝𝑙𝑖𝑒𝑠 𝒙𝒊 ≥ 𝒚𝒊 𝑓𝑜𝑟 𝒊
𝑎𝑛𝑑 𝒙𝒓 > 𝒚𝒓 𝑓𝑜𝑟 𝑎𝑡 𝑙𝑒𝑎𝑠𝑡 𝑜𝑛𝑒 𝒊 = 𝒓 𝑎𝑛𝑑 𝒙 > 𝒚 𝑖𝑚𝑝𝑙𝑖𝑒𝑠 𝒙𝒊 > 𝒚𝒊 𝑓𝑜𝑟 𝑒𝑎𝑐ℎ 𝑖 .

4. MOLFP with Rough Interval Coefficients

In this section, Multi-objective linear fractional programming problem with rough interval
coefficients (MOLFPRIC) is considered. Formulating an MOLFP model requires that Crisp
values be selected for the model coefficients. The values of several of these coefficients are
only approximately known. The major advantage of the proposed operations over the existing
one is that algorithm deal with uncertainty coefficients which take the form of fully rough
interval coefficients. Now, joining all the data in the MOLFP model is required.

Let us consider an MOLFPRIC as:

∑𝑛𝑗=1([𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] , [𝑐𝑖𝑗 𝑢
, 𝑐𝑖𝑗 𝑙
]) 𝑥𝑗 + ([𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 𝑙
] , [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ])
max 𝑧1 =
∑𝑛𝑗=1([𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 𝑙
] , [𝑑𝑖𝑗 𝑢
, 𝑑𝑖𝑗 ]) 𝑥𝑗 + ([𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ] , [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ])

∑𝑛𝑗=1([𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] , [𝑐𝑖𝑗 𝑢
, 𝑐𝑖𝑗 𝑙
]) 𝑥𝑗 + ([𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 𝑙
] , [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ])
max 𝑧2 =
∑𝑛𝑗=1([𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 𝑙
] , [𝑑𝑖𝑗 𝑢
, 𝑑𝑖𝑗 ]) 𝑥𝑗 + ([𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ] , [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ])
.
.
.

Volume 23, Issue 7, July - 2021 Page -97

 Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

.
∑𝑛𝑗=1([𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] , [𝑐𝑖𝑗 𝑢
, 𝑐𝑖𝑗 𝑙
]) 𝑥𝑗 + ([𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 𝑙
] , [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ])
max 𝑧𝑘 =
∑𝑛𝑗=1([𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 𝑙
] , [𝑑𝑖𝑗 𝑢
, 𝑑𝑖𝑗 ]) 𝑥𝑗 + ([𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ] , [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ])

S.T
𝑛
𝑙
∑([𝑎𝑖𝑗 𝑢
, 𝑎𝑖𝑗 𝑙
] , [𝑎𝑖𝑗 𝑢
, 𝑎𝑖𝑗 ]) 𝑥𝑗 ≤ [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ], [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ] 𝑖 = 1, 2, … , 𝑚
𝑗=1

𝑥𝑗 ≥ 0 𝑗 = 1, 2, … , 𝑛 (2)

where
𝒍 𝒖 𝒍 𝒖
([𝒄𝒍𝒊𝒋 , 𝒄𝒖𝒊𝒋 ] , [𝒄𝒍𝒊𝒋 , 𝒄𝒖𝒊𝒋 ]), ([𝒅𝒊𝒋 , 𝒅𝒊𝒋 ], [𝒅𝒊𝒋 , 𝒅𝒊𝒋 ]) are rough intervals coefficients of the objective
function (𝑗 = 1, 2, … , 𝑛 , 𝑖 = 1, 2, … , 𝑚) ,
𝑙 𝑢 𝑙 𝑢
([𝛼𝑖𝑗 , 𝛼𝑖𝑗 ] , [𝛼𝑖𝑗 , 𝛼𝑖𝑗 ]) are rough intervals constants of the numerator,

([𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ] , [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ]) are rough intervals constants of the denominator,

[𝑎𝑖𝑗 𝑙 , 𝑎𝑖𝑗 𝑢 ], [𝑎𝑖𝑗 𝑙 , 𝑎𝑖𝑗 𝑢 ] , [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ], [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ] are rough interval coefficients of the constraints.

𝑥 = (𝑥1 , 𝑥2 , … , 𝑥𝑛 )𝑇 denote the vector of all decision variables

5. Our Proposed Approach for Solving MOLFPRIC

𝑝𝑙 𝑝𝑢 𝑝𝑙 𝑝𝑢
We first find the possibly optimal range [𝑧1 , 𝑧1 ], [𝑧2 , 𝑧2 ] by solving MOLFP with the
interval coefficients problem. Secondly, we find the surly optimal range
[𝑧1 𝑠𝑙 , 𝑧1 𝑠𝑢 ] , [𝑧2 𝑠𝑙 , 𝑧2 𝑠𝑢 ] by solving MOLFP with the interval coefficients problem. Then we
transform MOLFP problems into linear programming problems using Guzel’s proposal [18].
Finally, the linear programming problems are solved by simplex method, whose optimal
solution is the required efficient solution of the original problem.

Input: Consider an MOLFPRIC problem illustrated in the model (2)

𝑝𝑙 𝑝𝑢 𝑝𝑙 𝑝𝑢
Step1: Find the possibly optimal range [𝑧1 , 𝑧1 ], [𝑧2 , 𝑧2 ] by solving the following
MOLFP with interval coefficients problem [16]:

∑𝑛𝑗=1[𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] 𝑥𝑗 + [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ]
max 𝑧1 =
∑𝑛𝑗=1[𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 ] 𝑥𝑗 + [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ]

Volume 23, Issue 7, July - 2021 Page -98

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

∑𝑛𝑗=1[𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] 𝑥𝑗 + [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ]
max 𝑧2 =
∑𝑛𝑗=1[𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 ] 𝑥𝑗 + [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ]

S.T
𝑛
𝑙
∑([𝑎𝑖𝑗 𝑢
, 𝑎𝑖𝑗 𝑙
] , [𝑎𝑖𝑗 𝑢
, 𝑎𝑖𝑗 ]) 𝑥𝑗 ≤ [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ], [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ] 𝑖 = 1, 2, … , 𝑚
𝑗=1

𝑥𝑗 ≥ 0 𝑗 = 1, 2, … , 𝑛 (3)

If the problem (3) is infeasible go to step 3.

Step 2: Find the surely optimal range [𝑧1 𝑠𝑙 , 𝑧1 𝑠𝑢 ] , [𝑧2 𝑠𝑙 , 𝑧2 𝑠𝑢 ] by solving the following
MODNP with interval coefficients problem [16]:

∑𝑛𝑗=1 [𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] 𝑥𝑗 + [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ]
max 𝑧1 =
∑𝑛𝑗=1 [𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 ] 𝑥𝑗 + [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ]

∑𝑛𝑗=1 [𝑐𝑖𝑗
𝑙 𝑢
, 𝑐𝑖𝑗 𝑙
] 𝑥𝑗 + [𝛼𝑖𝑗 𝑢
, 𝛼𝑖𝑗 ]
max 𝑧2 =
∑𝑛𝑗=1 [𝑑𝑖𝑗
𝑙 𝑢
, 𝑑𝑖𝑗 ] 𝑥𝑗 + [𝛽𝑖𝑗𝑙 , 𝛽𝑖𝑗𝑢 ]

S.T
𝑛
𝑙
∑([𝑎𝑖𝑗 𝑢
, 𝑎𝑖𝑗 𝑙
] , [𝑎𝑖𝑗 𝑢
, 𝑎𝑖𝑗 ]) 𝑥𝑗 ≤ [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ], [𝑏𝑖 𝑙 , 𝑏𝑖 𝑢 ] 𝑖 = 1, 2, … , 𝑚
𝑗=1

𝑥𝑗 ≥ 0 𝑗 = 1, 2, … , 𝑛 (4)

Step 3: Transform MOLFP with interval coefficients (possibly optimal range)

𝑝𝑙 𝑝𝑢 𝑝𝑙 𝑝𝑢
[𝑧1 , 𝑧1 ], [𝑧2 , 𝑧2 ] to two models where their feasible set are 𝑈𝐿 , 𝑈𝑈 , respectively.

Step 4: Transform MOLFP with interval coefficients (surly optimal range)

[𝑧1 𝑠𝑙 , 𝑧1 𝑠𝑢 ] , [𝑧2 𝑠𝑙 , 𝑧2 𝑠𝑢 ] to two models where their feasible set are 𝑈 𝐿 , 𝑈 𝑈 respectively.

𝑝𝑢 𝑝𝑢
Step 5: Solve the problem 𝑈𝑈 , and obtain the upper bound [𝑧1 , 𝑧2 ] by using Guzel’s
proposal.
𝑝𝑙 𝑝𝑙
Step 6: Solve the problem 𝑈𝐿 , and obtain the lower bound [𝑧1 , 𝑧2 ] by using Guzel’s
proposal.

Step 7: Solve the problem 𝑈 𝑈 and obtain the upper bound [𝑧1 𝑠𝑢 , 𝑧2 𝑠𝑢 ] by using Guzel’s
proposal.

Volume 23, Issue 7, July - 2021 Page -99

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

Step 8: Solve the problem 𝑈 𝐿 and obtain the lower bound [𝑧1 𝑠𝑙 , 𝑧2 𝑠𝑙 ] by using Guzel’s
proposal.
Step 9: There are three potential outcomes for MOLFPRIC Problem (2) as follows:
1) If MOLFP with interval coefficients Problems (3) and (4) have optimal ranges, then
MOLFPRIC Problem (2) has a rough range as
𝑝𝑙 𝑝𝑢 𝑝𝑙 𝑝𝑢
( [𝑧1 𝑠𝑙 , 𝑧1 𝑠𝑢 ] , [𝑧1 , 𝑧1 ]), ([𝑧2 𝑠𝑙 , 𝑧2 𝑠𝑢 ] , [𝑧2 , 𝑧2 ]).

2) If MOLFP with interval coefficients Problem (4) has unbounded range, then
MOLFPRIC Problem (2) has unbounded range.
3) If MOLFP with interval coefficients Problem (3) is infeasible, then MOLFPRIC
Problem (2) is infeasible.

Volume 23, Issue 7, July - 2021 Page -100

 Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

Start

Construct the mathematical model of MOLFP

with rough interval coefficients

Step 1: Find the possibly optimal range

𝑝𝑙 𝑝𝑢 𝑝𝑙 𝑝𝑢
[𝑧1 , 𝑧1 ], [𝑧2 , 𝑧2 ]

Feasible Infeasibl
e
Step 2: Find the surely optimal
range [𝑧1 𝑠𝑙 , 𝑧1 𝑠𝑢 ] , [𝑧2 𝑠𝑙 , 𝑧2 𝑠𝑢 ]

Step 3: Transform MOLFP with interval coefficients

𝑝𝑙 𝑝𝑢 𝑝𝑙 𝑝𝑢
(possibly optimal range) [𝑧1 , 𝑧1 ], [𝑧2 , 𝑧2 ] to
two models where their feasible set are 𝑈𝐿 , 𝑈𝑈 ,
respectively.

Step 4: Transform MOLFP with interval coefficients

(surly optimal range) [𝑧1 𝑠𝑙 , 𝑧1 𝑠𝑢 ] , [𝑧2 𝑠𝑙 , 𝑧2 𝑠𝑢 ] to two
models where their feasible set are
𝑈 𝐿 , 𝑈 𝑈 respectively.

Step5: Solve the problem Step6: Solve the problem Step 7: Solve the problem
𝑈𝑈 , and obtain the upper 𝑈𝐿 , and obtain the lower 𝑈 𝑈 and obtain the upper
𝑝𝑢 𝑝𝑢 𝑝𝑙 𝑝𝑙
bound [𝑧1 , 𝑧2 ] bound [𝑧1 , 𝑧2 ] bound [𝑧1 𝑠𝑢 , 𝑧2 𝑠𝑢 ]

If MOLFP with interval coefficients Probles

(step5:step8) have optimal ranges, then MOLFPRIC Step 8: Solve the problem 𝑈 𝐿 and
Problem has a rough range as
𝑠𝑙 𝑠𝑢 𝑝𝑙 𝑝𝑢 𝑠𝑙 𝑠𝑢 𝑝𝑙 𝑝𝑢 obtain the lower bound [𝑧1 𝑠𝑙 , 𝑧2 𝑠𝑙 ]
( [𝑧1 , 𝑧1 ] , [𝑧1 , 𝑧1 ]), ([𝑧2 , 𝑧2 ] , [𝑧2 , 𝑧2 ]).

stop

Figure 1: Flow chart to solve MOLFPRIC problems

Volume 23, Issue 7, July - 2021 Page -101

 Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

6. Numerical Example
Consider the following multi-objective linear fractional programming problem with rough
interval coefficients. (we modified This example in [10] with new assumptions)

[−3, −1][−4, −1]𝑥1 + [1,3][0.5,4]𝑥2 + [1,2][0.5,3]

min 𝑧1 (𝑥) =
[0.5,1][0.25,1.5]𝑥1 + [1,2][0.5,2.5]𝑥2 + [0.5,1][0.25,1.25]
[2,5][1,6]𝑥1 + [1,2][0.5,3]𝑥2 + [1,2][0.5,3]
min 𝑧2 (𝑥) =
[1,2][0.5,2.5]𝑥1 + [1,3][0.5,4]𝑥2 + [0.5,1][0.25,1.25]

s.t

[1,2][0.5,2.5]𝑥1 + [0.5,1][0.25,1.5]𝑥2 ≤ [2,4][1,5]

[2,3][1,4]𝑥1 − [3,2][4,1]𝑥2 ≤ [2,5][1,6] (5)

[0.5,1][0.25,1.75]𝑥1 + [1,2][0.5,2.5]𝑥2 ≤ [1,3][0.5,3.5]

[0.5,1][0.25,1.5]𝑥1 + [2,3][1,4]𝑥2 ≤ [1,2][0.5,3.5]

𝑥1 , 𝑥2 ≥ 0
To solve Problem (5), we have to solve two MOLFP with interval coefficients problems as
follows:

Step1:

[−4, −1]𝑥1 + [0.5,4]𝑥2 + [0.5,3]

min 𝑧1 (𝑥) =
[0.25,1.5]𝑥1 + [0.5,2.5]𝑥2 + [0.25,1.25]
[1,6]𝑥1 + [0.5,3]𝑥2 + [0.5,3]
min 𝑧2 (𝑥) =
[0.5,2.5]𝑥1 + [0.5,4]𝑥2 + [0.25,1.25]

s.t

[0.5,2.5]𝑥1 + [0.25,1.5]𝑥2 ≤ [1,5]

[1,4]𝑥1 − [4,1]𝑥2 ≤ [1,6] (6)

[0.25,1.75]𝑥1 + [0.5,2.5]𝑥2 ≤ [0.5,3.5]

[0.25,1.5]𝑥1 + [1,4]𝑥2 ≤ [0.5,3.5]

𝑥1 , 𝑥2 ≥ 0

Volume 23, Issue 7, July - 2021 Page -102

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

Step 2:

[−3, −1]𝑥1 + [1,3]𝑥2 + [1,2]

min 𝑓1 (𝑥) =
[0.5,1]𝑥1 + [1,2]𝑥2 + [0.5,1]
[2,5]𝑥1 + [1,2]𝑥2 + [1,2]
min 𝑓2 (𝑥) =
[1,2]𝑥1 + [1,3]𝑥2 + [0.5,1]

s.t

[1,2]𝑥1 + [0.5,1]𝑥2 ≤ [2,4]

[2,3]𝑥1 − [3,2]𝑥2 ≤ [2,5] (7)

[0.5,1]𝑥1 + [1,2]𝑥2 ≤ [1,3]

[0.5,1]𝑥1 + [2,3]𝑥2 ≤ [1,2]

𝑥1 , 𝑥2 ≥ 0
Step 3: The MOLFPIC Problem (6) is transformed to MOLFP problems R1 and R2, where
𝑙 𝑢
their feasible sets are 𝑈 𝑎𝑛𝑑 𝑈 , respectively.
Step4: The MOLFPIC Problem (7) is transformed to MOLFP problems R3 and R4, where
their feasible sets are 𝑈 𝑙 𝑎𝑛𝑑 𝑈 𝑢 , respectively.
𝑙 −4𝑥1 + 0.5𝑥2 +0.5 𝑢 −𝑥1 + 4𝑥2 +3
R1: 𝑧1 = min R2: 𝑧1 = min
0.25𝑥1 + 0.5𝑥2 +0.5 1.5𝑥1 + 2.5𝑥2 +1.25
𝑙 𝑥1 + 0.5𝑥2 +0.5 𝑢 6𝑥1 + 3𝑥2 +3
𝑧2 = min 𝑧2 = min
0.5𝑥1 + 0.5𝑥2 +0.25 2.5𝑥1 + 4𝑥2 +1.25
s.t s.t
2.5𝑥1 + 1.5𝑥2 ≤ 1 0.5𝑥1 + 0.25𝑥2 ≤ 5
4𝑥1 − 𝑥2 ≤ 1 𝑥1 − 4𝑥2 ≤ 6
1.75𝑥1 + 2.5𝑥2 ≤ 0.5 0.25𝑥1 + 0.5𝑥2 ≤ 3.5
1.5𝑥1 + 4𝑥2 ≥ 0.5 0.25𝑥1 + 𝑥2 ≥ 3.5
𝑥1 , 𝑥2 ≥ 0 𝑥1 , 𝑥2 ≥ 0
−3𝑥 + 𝑥 +1 −𝑥 + 3𝑥 +2
R3: 𝑧1 𝑙 = min 0.5𝑥 1+ 𝑥 2+0.5 R4: 𝑧1 𝑢 = min 𝑥 1+ 2𝑥 2+1
1 2 1 2
𝑙 2𝑥1 + 𝑥2 +1 𝑢 5𝑥1 + 2𝑥2 +2
𝑧2 = min 𝑧2 = min
𝑥1 + 𝑥2 +0.5 2𝑥1 + 3𝑥2 +1
s.t s.t
2𝑥1 + 𝑥2 ≤ 2 𝑥1 + 0.5𝑥2 ≤ 4
𝑥1 − 2𝑥2 ≤ 2 2𝑥1 − 3𝑥2 ≤ 5
𝑥1 + 2𝑥2 ≤ 1 0.5𝑥1 + 𝑥2 ≤ 3
𝑥1 + 3𝑥2 ≥ 1 0.5𝑥1 + 2𝑥2 ≥ 2
𝑥1 , 𝑥2 ≥ 0 𝑥1 , 𝑥2 ≥ 0

Volume 23, Issue 7, July - 2021 Page -103

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

𝑝𝑢 𝑝𝑢
Step 5: Solve the problem 𝑈𝑈 , and obtain the upper bound [𝑧1 , 𝑧2 ] (R2) by using
Guzel’s proposal.
𝑢 −𝑥1 + 4𝑥2 +3
𝑧1 = min 1.5𝑥1 + 2.5𝑥2 +1.25

𝑢 6𝑥1 + 3𝑥2 + 3
𝑧2 = min
2.5𝑥1 + 4𝑥2 + 1.25

s.t

0.5𝑥1 + 0.25𝑥2 ≤ 5

𝑥1 − 4𝑥2 ≤ 6

0.25𝑥1 + 0.5𝑥2 ≤ 3.5

0.25𝑥1 + 𝑥2 ≥ 3.5

𝑥1 , 𝑥2 ≥ 0
To solve this MOLFP problem, we find the optimal value of each of the objective functions
𝑢 𝑢
𝑧1 , 𝑧2 subject to the above constraints, using any of the methods for solving linear
fractional programming problems (we used the equivalent of Charnes A. and Cooper W.W. [2]).
We get,
𝑢
min 𝑧1 = −0.10
𝑢
min 𝑧2 = 0.82
An LP problem, which is equivalent to the MOLFP problem, is constructed according to the
proposed algorithm as follows:

min 𝑈𝑈 = (−𝑥1 + 4𝑥2 + 3) + (0.10)(1.5𝑥1 + 2.5𝑥2 + 1.25) + (6𝑥1 + 3𝑥2 + 3)

− (0.82)(2.5𝑥1 + 4𝑥2 + 1.25)

= 3.1𝑥1 + 3.97𝑥2 + 5.1

s.t

0.5𝑥1 + 0.25𝑥2 ≤ 5

𝑥1 − 4𝑥2 ≤ 6

0.25𝑥1 + 0.5𝑥2 ≤ 3.5

0.25𝑥1 + 𝑥2 ≥ 3.5

𝑥1 , 𝑥2 ≥ 0
𝑢 𝑢 𝑢 𝑢
the resulting 𝑥1𝑜𝑝𝑡 , 𝑥2𝑜𝑝𝑡 = (0, 3.5) and 𝑧1𝑜𝑝𝑡 , 𝑧2𝑜𝑝𝑡 = (1.7, 0.89)

Volume 23, Issue 7, July - 2021 Page -104

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

𝑝𝑙 𝑝𝑙
Step 6: Solve the problem 𝑈𝐿 , and obtain the lower bound [𝑧1 , 𝑧2 ] (R1) by using Guzel’s
proposal.
𝑙 −4𝑥1 + 0.5𝑥2 +0.5
𝑧1 = min 0.25𝑥1 + 0.5𝑥2 +0.5

𝑙 𝑥1 + 0.5𝑥2 + 0.5
𝑧2 = min
0.5𝑥1 + 0.5𝑥2 + 0.25

s.t

2.5𝑥1 + 1.5𝑥2 ≤ 1

4𝑥1 − 𝑥2 ≤ 1

1.75𝑥1 + 2.5𝑥2 ≤ 0.5

1.5𝑥1 + 4𝑥2 ≥ 0.5

𝑥1 , 𝑥2 ≥ 0
𝑙
min 𝑧1 = −1.45
𝑙
min 𝑧2 = 1.71
An LP problem, which is equivalent to the MOLFP problem, is constructed according to the
proposed algorithm as follows:

min 𝑈𝐿 = (−4𝑥1 + 0.5𝑥2 + 0.5) + (1.45)(0.25𝑥1 + 0.5𝑥2 + 0.5) + (𝑥1 + 0.5𝑥2 + 0.5)
− (1.71)(0.5𝑥1 + 0.5𝑥2 + 0.25)

= −3.5𝑥1 + 0.9𝑥2 + 0.93

s.t

2.5𝑥1 + 1.5𝑥2 ≤ 1

4𝑥1 − 𝑥2 ≤ 1

1.75𝑥1 + 2.5𝑥2 ≤ 0.5

1.5𝑥1 + 4𝑥2 ≥ 0.5

𝑥1 , 𝑥2 ≥ 0
𝑙 𝑙 𝑙 𝑙
The resulting 𝑥1𝑜𝑝𝑡 , 𝑥2𝑜𝑝𝑡 = (0.23, 0.038) and 𝑧1𝑜𝑝𝑡 , 𝑧2𝑜𝑝𝑡 = (−0.7, 1.95)

Step 7: Solve the problem 𝑈 𝑈 and obtain the upper bound [𝑧1 𝑠𝑢 , 𝑧2 𝑠𝑢 ] (R4) by using Guzel’s
proposal.
−𝑥1 + 3𝑥2 + 2
𝑧1 𝑢 = min
𝑥1 + 2𝑥2 + 1

Volume 23, Issue 7, July - 2021 Page -105

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

5𝑥1 + 2𝑥2 + 2
𝑧2 𝑢 = min
2𝑥1 + 3𝑥2 + 1

s.t

𝑥1 + 0.5𝑥2 ≤ 4

2𝑥1 − 3𝑥2 ≤ 5

0.5𝑥1 + 𝑥2 ≤ 3

0.5𝑥1 + 2𝑥2 ≥ 2

𝑥1 , 𝑥2 ≥ 0

min 𝑧1 𝑢 = −0.02
min 𝑧2 𝑢 = 0.8

An LP problem, which is equivalent to the MOLFP problem, is constructed according to the

proposed algorithm as follows:

min 𝑈 𝑈 = (−𝑥1 + 3𝑥2 + 2) + (0.02)(𝑥1 + 2𝑥2 + 1) + (5𝑥1 + 2𝑥2 + 2)

− (0.8)(2𝑥1 + 3𝑥2 + 1)

= 2.42𝑥1 + 2.64𝑥2 + 3.22

s.t

𝑥1 + 0.5𝑥2 ≤ 4

2𝑥1 − 3𝑥2 ≤ 5

0.5𝑥1 + 𝑥2 ≤ 3

0.5𝑥1 + 2𝑥2 ≥ 2

𝑥1 , 𝑥2 ≥ 0

the resulting 𝑥1𝑜𝑝𝑡 𝑢 , 𝑥2𝑜𝑝𝑡 𝑢 = (0, 1) and 𝑧1𝑜𝑝𝑡 𝑢 , 𝑧2𝑜𝑝𝑡 𝑢 = (1.67, 1)

Step 8: Solve the problem 𝑈 𝐿 and obtain the lower bound [𝑧1 𝑠𝑙 , 𝑧2 𝑠𝑙 ] (R3) by using Guzel’s
proposal.
−3𝑥1 + 𝑥2 + 1
𝑧1 𝑙 = min
0.5𝑥1 + 𝑥2 + 0.5

2𝑥1 + 𝑥2 + 1
𝑧2 𝑙 = min
𝑥1 + 𝑥2 + 0.5

s.t

2𝑥1 + 𝑥2 ≤ 2

Volume 23, Issue 7, July - 2021 Page -106

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

𝑥1 − 2𝑥2 ≤ 2

𝑥1 + 2𝑥2 ≤ 1

𝑥1 + 3𝑥2 ≥ 1

𝑥1 , 𝑥2 ≥ 0

min 𝑧1 𝑙 = −1.17
min 𝑧2 𝑙 = 1.5

An LP problem, which is equivalent to the MOLFP problem, is constructed according to the

proposed algorithm as follows:

min 𝑈 𝐿 = (−3𝑥1 + 𝑥2 + 1) + (1.17)(0.5𝑥1 + 𝑥2 + 0.5) + (2𝑥1 + 𝑥2 + 1)

− (1.5)(𝑥1 + 𝑥2 + 0.5)

= −1.91𝑥1 + 1.67𝑥2 + 1.84

s.t

2𝑥1 + 𝑥2 ≤ 2

𝑥1 − 2𝑥2 ≤ 2

𝑥1 + 2𝑥2 ≤ 1

𝑥1 + 3𝑥2 ≥ 1

𝑥1 , 𝑥2 ≥ 0

the resulting 𝑥1𝑜𝑝𝑡 𝑙 , 𝑥2𝑜𝑝𝑡 𝑙 = (0.727, 0.090) and 𝑧1𝑜𝑝𝑡 𝑙 , 𝑧2𝑜𝑝𝑡 𝑙 = (−1.14, 1.93)

The optimal values and the optimal solutions of MOLFP (R1: R4) are given in Table 1.
Table 1: Optimal values and optimal solutions of numerical example
Problem R1 R2 R3 R4
Optimal values (−0.7, 1.95) (1.7, 0.89) (-1.14, 1.93) (1.67, 1)
𝑢 𝑢
(𝒛𝟏 , 𝒛𝟐 ) 𝑙 𝑙
(𝑧1 , 𝑧2 ) (𝑧1 , 𝑧2 ) (𝑧1 𝑙 , 𝑧2 𝑙 ) (𝑧1 𝑢 , 𝑧2 𝑢 )
Optimal values with 0.159 19 0.603 5.86
Single obj. 𝑈𝑙 𝑈𝑢 𝑈𝑙 𝑈𝑢
(𝑼𝒍 , 𝑼𝒖 )
Optimal solutions (0.23, 0.038) (0, 3.5) (0.727, 0.090) (0, 1)
𝑢 𝑢
(𝒙𝟏 , 𝒙𝟐 ) 𝑙 𝑙
(𝑥1 , 𝑥2 ) (𝑥1 , 𝑥2 ) (𝑥1 𝑙 , 𝑥2 𝑙 ) (𝑥1 𝑢 , 𝑥2 𝑢 )

So, for problem (2):

- [𝑈 𝑙 , 𝑈 𝑢 ] = [0.603, 5.86] is the surely optimal range,

- [𝑈𝑙 , 𝑈 𝑢 ] = [0.159, 19] is the possibly optimal range,

Volume 23, Issue 7, July - 2021 Page -107

 Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

- ([𝑈 𝑙 , 𝑈 𝑢 ] [𝑈𝑙 , 𝑈 𝑢 ]) = ([0.603, 5.86] [0.159, 19]) is the rough optimal range,
- The solutions (0.23, 0.038)𝑡 𝑎𝑛𝑑 (0, 3.5)𝑡 are two of the rather satisfactory solutions,
- The solutions (0.727, 0.090)𝑡 𝑎𝑛𝑑 (0, 1)𝑡 are two of the completely satisfactory
solutions.

7. Conclusion
In this paper, Basic concepts of rough intervals are studied. Rough intervals are valuable and
novel tools to process the uncertainty in decision making problems with fractional
programming. A new assumption of MOLFP problems is presented in which all of the
coefficients. In order to solve these types of problems, we presented that each one of them can
be transformed into two MOLFP problems with interval coefficients. A numerical example is
given to illustrate our motivation for considering MOLFP problems with rough intervals. That
algorithm as a methodology for this problem has solved it successfully after converted the
MOLFP problem into a single objective linear programming problem. Studying more
properties of the presented models can be a subject for further research.

References
[1] A. N. Dheyab, Finding the optimal solution for fractional linear programming problems
with fuzzy numbers. Journal of Kerbala University, 10(3) (2012), pp. 105-110.
[2] A. Charnes and W.W. Cooper, Programming with Linear Fractional Functionals, Naval
Research Logistics Quarterly 9 (1962), pp. 181-186.
[3] B. Liu, Theory and Practice of Uncertain Programming, Physica Verlag, New York,
(2002).

[4] B. Martos and V. Whinston, Hyperbolic programming. Naval Research Logistics (NRL),
11(2) (1964), pp. 135-155.
[5] E. Fathy, Building fuzzy approach with linearization technique for fully rough multi-
objective multi-level linear fractional programming problem, Iranian Journal of Fuzzy
Systems, Vol. 18, No. 2, (2021), pp. 139-157.
[6] E. S. ammar and M. muamer, On Solving Fuzzy Rough Linear Fractional Programming
Problem, International Research Journal of Engineering and Technology, Vol. 03, Issue 04,
(2016), pp. 2099-2120.

[7] E. S. ammar and M. muamer, Algorithm for Solving Multi Objective Linear Fractional
Programming Problem with Fuzzy Rough Coefficients, Open Science Journal of Mathematics
and Application, Vol. 04, Issue 01, (2016), pp. 1-8.
[8] E. A. Youness, Characterizing solutions of rough programming problems. European
Journal of Operational Research, vol. 168(3) (2006), pp. 1019–1029.
[9] E. B. Bajalinov, Linear-fractional programming: theory, methods, applications and
software. Kluwer Academic Publishers, Boston, (2003).

Volume 23, Issue 7, July - 2021 Page -108

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

[10] F. A. Pramy, An Approach for Solving Fuzzy Multi-Objective Linear Fractional

Programming Problems, International Journal of Mathematical, Engineering and Management
Sciences, Vol. 3, No. 3, (2018), pp. 280–293.

[11] J. Nasiri and M. Mashinchi, Rough set and data analysis in decision tables, Journal of
Uncertain Systems, vol. 3(3) (2009), pp. 232–240.
[12] J. Krysinski, Rough sets in the analysis of the structure-activity relationships of
antifungal imidazolium compounds, Journal of Pharmaceutical Sciences, vol. 84(2) (1995),
pp. 243–248.
[13] K. Swarup, Linear fractional functionals programming. Operations Research, 13(6)
(1965), pp. 1029-1036.
[14] M. S. Osman, K. R. Raslan, O. E. Emam and F. A. Farahat, Solving Multi-level Multi-
objective Fractional Programming Problem with Rough Intervals in the Objective Functions,
British Journal of Mathematics & Computer Science, vol. 21(2), (2017), pp. 1-17.
[15] M. S. Osman, E.F. Lashein, E.A. Youness and T.E.M. Atteya, Mathematical
programming in rough environment Optimization. vol. 60(5) (2011), pp. 603–611.
[16] M. Rebolledo, Rough intervals-enhancing intervals for qualitative modeling of technical
systems. Artificial Intelligence, vol. 170 (2006), pp. 667–685.
[17] M. Arabani and M.A.L. Nashaei, Application of Rough Set theory as a new approach to
simplify dams location, Scientia Iranica, vol. 13(2) (2006), pp. 152–158.
[18] N. Guzel, A proposal to the solution of multiobjective linear fractional programming
problem. Abstract and Applied Analysis (Vol. 2013). Hindawi Publishing Corporation,
(2013).

[19] N. Guzel & M. Sivri, Proposal of a solution to multi objective linear fractional
programming problem. Sigma Journal of Engineering and Natural Sciences, vol 2, (2005),
pp. 43–50.
[20] R. S. Porchelvi, L. Vasanthi, & R. I. Hepzibah,. On solving multi objective fractional
linear programming problems. International Journal of Current Research, 6(8) (2014), pp.
8095-8102.
[21] R. Moore, Interval Analysis, Prentice-Hall, Englewood-Cliffs, NJ, (1966).
[22] S. Jain,. Modeling of Gauss elimination technique for multi-objective linear
programming problem. Journal of Novel Applied Sciences (2012), 1, pp. 25-29.

[23] S. Jain,. Modeling of Gauss elimination technique for multi-objective fractional

programming problem. South Asian Journal of Mathematics, 4(3) (2014), pp. 148-153.
[24] W. Dinkelbach, On nonlinear fractional programming. Management Science, 13(7)
(1967), pp. 492-498.
[25] Y. Weiguo, L. Mingyu and L. Zhi, Variable precision rough set based decision tree
classifier, Journal of Intelligent and Fuzzy Systems, vol. 23(2) (2012), pp. 61–70.

Volume 23, Issue 7, July - 2021 Page -109