See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/368485402

A new technique to solve the assignment problems

Conference Paper in AIP Conference Proceedings · February 2023

DOI: 10.1063/5.0114724

CITATIONS READS
8 500

2 authors:

Zahraa K. Hashim Mushtak A.K. Shiker

4 PUBLICATIONS 23 CITATIONS
University of Babylon
79 PUBLICATIONS 1,437 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Mushtak A.K. Shiker on 08 August 2024.

The user has requested enhancement of the downloaded file.

A new technique to solve the assignment
problems
Cite as: AIP Conference Proceedings 2414, 040018 (2023); https://doi.org/10.1063/5.0114724
Published Online: 13 February 2023

Zahraa K. Hashim and Mushtak A. K. Shiker

AIP Conference Proceedings 2414, 040018 (2023); https://doi.org/10.1063/5.0114724 2414, 040018

© 2023 Author(s).
 A New Technique to Solve the Assignment Problems
Zahraa K. Hashim 1, a) and Mushtak A. K. Shiker 1, b)

1
Department of mathematics, collage of education for pure sciences, university of Babylon, Babil- Iraq
a)
[email protected]
b) Corresponding author: [email protected] , [email protected]

Abstract. The objective of the assignment problem is to assign number of jobs to an equal number of persons, it's an
important part of transportation problem where it is employed to maximize the total profit or to minimize the total cost of
allocation. In this work, a new modification of the assignment problems is introduced to find the optimal solution for
these problems, we compare this solution with the solution of Hungarian method, they are equal in all examples under
study.

Keywords: Assignment problem. Reducing cost. Reducing time. Hungarian method

INTRODUCTION

Assignment problem (AP) is one of linear programming problems, it is a special case of a transportation problem
[1]. (AP) aims to reduce the cost or the time (or both) to complete a specified job by assigning mangers to projects,
assigning business to machines, assigning the employees to complete the jobs, assigning the means to transport
goods form one place to another and distributing contracts to contractors. (AP) works under the following principles
[2]:
1. Equal number of jobs to employees.
2. Every employee accomplishes one job.
3. Equal workers to work on specific machines.
4. Non negative values.
5. The cost of performing the job is known and predetermined.
There are many methods are used to solve the (AP) such as the combinations method, linear programming
method, and the Hungarian method which is the common method [3]. The Hungarian method went through several
developments by many of scientists. In 1955, Harold Kuhn [4, 5] was developed and published it and named it as
(Hungarian method) because it discovered by two Hungarian mathematicians Denes Kuhn and Jeno Egervary. Since
1957, the Hungarian method also known as Kuhn Munkres algorithm or Munkres assignment algorithm because
James Munkres [6] reviewed it and observed that it is strongly polynomial. In 2020 [7- 9], Hussein H A and Shiker
M A K found the optimal solution of the assignment problems by using two new effective methods. The authors
worked in various fields to find the optimal solution by using conjugate gradient methods [10- 13], by using trust
region method [14- 18], by using projection method [19- 24], by using line search techniques [25, 26], and so on, see
[27- 32], but in this work we suggest a modification method to solve the assignment problems.

A New Algorithm (Az1)

The new suggested algorithm is using to find the optimal solution of (AP) depending on several steps for the
balanced and unbalanced assignment problems.
• In the case of balanced assignment problem, (AZ1) algorithm works depending on the following steps:
1. Subtracts the lowest element in each row from all elements in that row.
2. In every row and in every column it must contain at least one zero.

The Sixth Local Scientific Conference-The Third Scientific International

AIP Conf. Proc. 2414, 040018-1–040018-7; https://doi.org/10.1063/5.0114724
Published by AIP Publishing. 978-0-7354-4431-7/$30.00

040018-1
3. Determines the rows and columns that contains a unique zero and chooses these zeros and deletes these rows and
columns respectively.
4. If there are more than one zero in any column, we apply the rules |𝑥 − 𝑟|, |𝑥 − 𝑙| for each zero, where 𝑥, 𝑟, and 𝑙
represent the lowest value, the right of 𝑥, and the left of 𝑥 (in the origin table) respectively, (but if there is no
element on the left or the right of 𝑥 we apply the difference for the other side), then we choose the cell that give the
largest value and deletes the row and column that contain this cell.
5. For the remaining row without zeros, we subtract the lowest element in that row from all elements of the row
except the zeros, and continue with the same steps.
6. Now we can find the optimal solution (for example, total cost or total time) from all assignment cells (zeros).
• In the case of unbalanced assignment problem, (AZ1) algorithm works depending on the following steps::
1. Adds an imaginary row when the number of columns > number of rows or an imaginary column when the
number of rows > number of columns.
2. if an imaginary row added, the new algorithm works normally as well as balance case.
3. If an imaginary column added, then the new algorithm subtracts the smallest element of each column from that
column elements and then the method well work as well as balance case.

Numerical Examples
Example1. A company wants to assign four jobs to four groups so that the total completion time (in hour) is the
least possible. The required time for each job is ordered to be completed as shown in the table below.

TABLE 1. Time matrix

Order1 Order2 Order3 Order4

Group
Group 1 5 7 9 10
Group 2 12 8 5 6
Group 3 6 9 11 9
Group 4 7 13 8 6

Solution: To solve this problem, we'll follow the (AZ1) algorithm. At first we have to look for the smallest element
in each row and subtract it from the rest of the row:

TABLE 2. Time matrix after subtracting the smallest element

Order1 Order2 Order3 Order4

Group
Group 1 0 2 4 5
Group 2 7 3 0 1
Group 3 0 3 5 3
Group 4 1 7 2 0
Note that the second column does not contain a zero. we subtract smallest element in the second column from the
column element.

TABLE 3. Time matrix after subtracting the least element of the second column

Order1 Order2 Order3 Order4

Group
Group 1 0 0 4 5
Group 2 7 1 0 1
Group 3 0 1 5 3
Group 4 1 5 2 0

040018-2
Note that there are two smallest elements in the first column (in the first and third rows), so, we apply the rule |𝑥 −
𝑟|, |𝑥 − 𝑙| → | 5 − 7 | = 2 , | 6 − 9 | = 3, we choose 6 and delete the third row and first column.

TABLE 4. Optimal assignment problem matrix

Order1 Order2 Order3 Order4

Group
Group 1 0 0 4 5
Group 2 7 1 0 1
Group 3 0 1 5 3
Group 4 1 5 2 0

So, the optimal solution is: T time = 7 + 6 + 5 + 6 → T = 24

Example 2. A contraction company intends to build four residential project in four different regions. If the company
has three means dig and level the land and the company’s estimate of the costs of completion is in the thousands of
dollars as in the following table :

TABLE 5. Cost matrix (in the thousands of dollars)

Region1 Region2 Region3 Region4

Means
A 9 12 8 11
B 16 5 18 9
C 7 4 9 20

If we want to find the best solution, we must first add an imaginary row D

TABLE 6. Cost matrix after adding an imaginary row

Region1 Region2 Region3 Region4

Means
A 9 12 8 11
B 16 5 18 9
C 7 4 9 20
D 0 0 0 0

We look for the smallest element in each row and subtract it from the rest of the row elements.

TABLE 7. Cost matrix after Subtracting the least element

Region1 Region2 Region3 Region4

Means
A 1 4 0 3
B 11 0 13 4
C 3 0 5 16
D 0 0 0 0

040018-3
 Because there are two zeros in the same column we apply the rule |𝑥 − 𝑟|, |𝑥 − 𝑙| → | 5 − 18 | = 13 , | 5 −
16 | = 11 and | 4 − 9 | = 5 , | 4 − 7 | = 3, we choose zero which is equal to 5 and delete the second row and
second column, and for the third row we take the smallest element and we subtract it from the rest of the row
elements except zeros.

TABLE 8. Optimal assignment problem matrix

Region1 Region2 Region3 Region4

Means
A 1 4 0 3
B 11 0 13 4
C 0 0 5 13
D 0 0 0 0

So, the optimal solution is: T = 5 + 7 + 8 → T = 20

Example 3. For the following unbalanced assignment problem:

TABLE 9. Time matrix

Job1 Job2 Job3 Job4 Job5

Worker
A 5 2 4 2 5
B 2 4 7 6 6
C 6 7 5 8 7
D 5 2 3 3 4
E 8 3 7 8 6
F 3 6 3 5 7

If we want to minimize the total time for performing all the jobs, we have to add an imaginary column 6.

TABLE 10. Time matrix after adding an imaginary column

Job1 Job2 Job3 Job4 Job5 Job6

Worker
A 5 2 4 2 5 0
B 2 4 7 6 6 0
C 6 7 5 8 7 0
D 5 2 3 3 4 0
E 8 3 7 8 6 0
F 3 6 3 5 7 0

We look for the smallest element in each column and subtract it from the rest of the column.

040018-4
 TABLE 11. Time matrix after Subtracting the least element

Job1 Job2 Job3 Job4 Job5 Job6

Worker
A 3 0 1 0 1 0
B 0 2 4 4 2 0
C 4 5 2 6 3 0
D 3 0 0 1 0 0
E 6 1 4 6 2 0
F 1 4 0 3 3 0

we choose the only zero from each column while deleting the row and column, because there are two zeros in the
same column, we apply the rule |x-l|→| 0-7 |=7 , and ,|0-3|=3 we choose zero which in column six and row third and
deleting the column and row, After selecting the remaining column without zeros, we subtract the least element in it
from the rest of the elements.

TABLE 12. Optimal assignment problem matrix

Job1 Job2 Job3 Job4 Job5 Job6

Worker
A 3 0 1 0 1 0
B 0 2 4 4 2 0
C 4 4 2 6 3 0
D 3 0 0 1 0 0
E 6 0 4 6 2 0
F 1 4 0 3 3 0

The optimal solution is determining from the zeros that highlighted by yellow color, while we neglect the zero with
red color because it lies in the imaginary column 6: T = 2 + 3 + 3 + 2 + 4 = 14

Comparison the Solutions

Here we compare the results of the suggested algorithm with the results of the Hungarian method in the following
table:

TABLE 13. Comparison between the results of Hungarian method and AZ1 method

Example1 Example2 Example3

Hungarian method 24 20 14
(AZ1) method 24 20 14

We can see that the results of the two algorithms are equal which indicated the efficiency and robustness of the new
algorithm. We run (AZ1) algorithm on many examples, and in all of them the optimal solutions of (AZ1) algorithm
are equal to the optimal solutions of the Hungarian method but in less steps, that is mean the (AZ1) algorithm is
more quickly than the Hungarian method with same optimal solutions.

040018-5
 CONCLUSION

In this work, the new suggested algorithm (AZ1) is introduced to find the optimal solution for assignment
problems. (AZ1) algorithm framework characterized by simplicity and ease, it is applying depending on several easy
steps. We applied (AZ1) algorithm on many examples to find the optimal solution, then we compared its results with
the results of the Hungarian method which is the classical historical method. The results of the two algorithms are
equal in all examples under study, but the new suggested method is more quickly and leas steps comparing with the
Hungarian method, that refers to the efficiency and robustness of the new algorithm.

REFERENCES

1. J. K. Strayer, Linear Programming and its Applications. Springer-Verlag, New York Inc, 1989.
2. S. H. Frederick and J. L. Gerald, Introduction to operation research. New York, NY: McGraw- Hill
Education. Nine edition, 2015.
3. M. S. Gaglani, A study on transportation Problem, transportation Problem, Assignment Problem and
Supply Chain Management. Thesis PhD, Saurashtra University, 2011.
4. W. K. Harold, The Hungarian method for assignment problem. Naval Research Logistics Quarterly, vol. 2,
pp. 83- 97, 1955.
5. W. K. Harold, Variants of the Hungarian method for assignment problems. Naval Research Logistics
Quartly, vol. 3, pp. 253- 258, 1956.
6. J. Munkres, Algorithm for the assignment and transportation problems, Journal of the Society for Industrial
and Applied Mathematics, vol. 5, no. 1, pp. 32- 38, 1957.
7. H. A. Hussein and M. A. K. Shiker, Two new effective methods to find the optimal solution for the
assignment problems, Journal of Advanced Research in Dynamical and Control Systems, vol.12, no. 7, pp.
49- 54, 2020.
8. H. A. Hussein, M. A. K. Shiker, and M. S. M. Zabiba, A new revised efficient of VAM to find the initial
solution for the transportation problem, J. Phys.: Conf. Ser. 1591 012032, 2020.
9. H. A. Hussein and M. A. K. Shiker, A modification to Vogel’s approximation method to solve
transportation problems, J. Phys.: Conf. Ser. 1591 012029, 2020.
10. H. A. Wasi and M. A. K. Shiker, Proposed CG method to solve unconstrained optimization problems, J.
Phys.: Conf. Ser. 1804 012024, 2021.
11. H. A. Wasi and M. A. K. Shiker, A new hybrid CGM for unconstrained optimization problems, J. Phys.:
Conf. Ser. 1664 012077, 2020.
12. M. M. Mahdi and M. A. K. Shiker, Three-Term of New Conjugate Gradient Projection Approach under
Wolfe Condition to Solve Unconstrained Optimization Problems, Journal of Advanced Research in
Dynamical and Control Systems, vol. 12, no. 7, pp. 788- 795, 2020.
13. H. A. Wasi and M. A. K. Shiker, Nonlinear conjugate gradient method with modified Armijo condition to
solve unconstrained optimization, J. Phys.: Conf. Ser. 1818 012021, 2021
14. M. A. K. Shiker and Z. Sahib A modified trust-region method for solving unconstrained optimization,
Journal of Engineering and Applied Sciences, vol. 13, no. 22, pp. 9667– 9671, 2018.
15. H. H. Dwail and M. A. K. Shiker Using a trust region method with nonmonotone technique to solve
unrestricted optimization problem, J. Phys.: Conf. Ser. 1664 012128, 2020.
16. H. H. Dwail et al., A new modified TR algorithm with adaptive radius to solve a nonlinear systems of
equations, J. Phys.: Conf. Ser. 1804 012108, 2021.
17. H. H. Dwail and M. A. K. Shiker, Reducing the time that TRM requires to solve systems of nonlinear
equations, IOP Conf. Ser.: Mater. Sci. Eng. 928 042043, 2020.
18. H. H. Dwail and M. A. K. Shiker Using trust region method with BFGS technique for solving nonlinear
systems of equations, J. Phys.: Conf. Ser. 1818 012022, 2021.
19. M. A. K. Shiker and K. Amini, A new projection-based algorithm for solving a large scale nonlinear
system of monotone equations, Croatian operational research review, vol. 9, no. 1, pp. 63- 73, 2018.
20. M. M. Mahdi and M. A. K. Shiker, A new projection technique for developing a Liu-Storey method to
solve nonlinear systems of monotone equations, J. Phys.: Conf. Ser. 1591 012030.
21. M. M. Mahdi and M. A. K. Shiker, Three terms of derivative free projection technique for solving
nonlinear monotone equations, J. Phys.: Conf. Ser. 1591 012031, 2020.

040018-6
 22. M. M. Mahdi and M. A. K. Shiker, Solving systems of nonlinear monotone equations by using a new
projection approach, J. Phys.: Conf. Ser. 1804 012107, 2021.
23. M. M. Mahdi and M. A. K. Shiker, A New Class of Three-Term Double Projection Approach for Solving
Nonlinear Monotone Equations J. Phys.: Conf. Ser. no. 1664 012147, 2020.
24. N. K. Dreeb, et al., Using a New Projection Approach to Find the Optimal Solution for Nonlinear Systems
of Monotone Equation, J. Phys.: Conf. Ser. 1818 012101, 2021.
25. K. H. Hashim and M. A. K. Shiker, Using a new line search method with gradient direction to solve
nonlinear systems of equations, J. Phys.: Conf. Ser. 1804 012106, 2021.
26. K. H. Hashim et al., Solving the Nonlinear Monotone Equations by Using a New Line Search Technique, J.
Phys.: Conf. Ser. 1818 012099, 2021.
27. Z. A. H. Hassan and E. K. Mutar, Geometry of reliability models of electrical system used inside
spacecraft, Second Al-Sadiq International Conference on Multidisciplinary in IT and Communication
Science and Applications (AIC-MITCSA), pp. 301-306, 2017
28. Z. A. H. Hassan and V. Balan, Fuzzy T-map estimates of complex circuit reliability, International
Conference on Current Research in Computer Science and Information Technology (ICCIT-2017), IEEE,
Special issue, pp. 136-139, 2017
29. Z. A. H. Hassan and V. Balan, Reliability extrema of a complex circuit on bi-variate slice classes, Karbala
International Journal of Modern Science, vol. 1, no. 1, pp. 1- 8, 2015
30. L. H. Hashim et al., An application comparison of two Poisson models on zero count data, J. Phys.: Conf.
Ser. 1818 012165, 2021.
31. H. A. Wasi and M. A. K. Shiker, A modified of FR method to solve unconstrained optimization, J. Phys.:
Conf. Ser. 1804 012023., 2021.
32. L. H. Hashim et al., An application comparison of two negative binomial models on rainfall count data, J.
Phys.: Conf. Ser. 1818 012100, 2021.

040018-7
View publication stats