EasyChair Preprint

№ 5915

Comparison of Performances of Jaya Algorithm

and Cuckoo Search Algorithm Using Benchmark
Functions

Mashuk Ahmed, Abdullah B. Nasser, Kamal Z. Zamli and

Sulistyo Heripracoyo

EasyChair preprints are intended for rapid

dissemination of research results and are
integrated with the rest of EasyChair.

June 27, 2021

 Comparison of Performances of Jaya Algorithm and
Cuckoo Search Algorithm Using Benchmark Functions

Mashuk Ahmed1, Abdullah B. Nasser1, Kamal Z. Zamli1, and Sulistyo Heripracoyo2

1Faculty of Computing, College of Computing and Applied Sciences, Universiti Malaysia
Pahang, , Pekan 26600, Pahang, Malaysia
2 School of Information Systems, Bina Nusantara University, Information Systems Department,
Jakarta, 11480, Indonesia
MCS20004@student.ump.edu.my
abdullahnasser@ump.edu.my
kamalz@ump.edu.my
hpracoyo@binas.edu

Abstract. Nowadays, selecting the best possible solution among several solu-
tions becomes an important skill for engineering and research. Therefore, engi-
neers are turning to optimization methods as a complementary alternative strat-
egy of exhaustive searching. Metaheuristic algorithms have been used success-
fully for solving different optimization problems. To help engineers select the
best metaheuristic algorithms for their problems, there is a need to evaluate the
performance of different metaheuristic algorithms against each other using
common case studies. This paper aims to compare the performance of two me-
taheuristic algorithms which are Jaya Algorithm (JA) and Cuckoo Search (CS)
using some common benchmark functions. CS and JA have implemented in the
same platform (Intellij IDEA Community Edition 2020.2.3) using the same lan-
guage (Java). The experimental results show that JA has better and consistent
performance as compared to CS in most cases in terms of execution time and
test suite size; however, the performance of JA is still within acceptable ranges.

Keywords: Jaya Algorithm, Cuckoo Search Algorithm, Metaheuristic Algo-

rithm, Optimization, Execution Time.

1 Introduction
Besides the analysis, optimization has been indicated as one of the vital stages of
engineering design [1]. Optimization is a vital part of software engineering which
allows the engineer to find the optimum solution in the presence of design constraints
and criteria. Whereas the analysis stage of the engineering design is the use of the
mathematic model to predict the design results optimization method such as metaheu-
ristic algorithms have been used successfully for solving different optimization prob-
lems such as robotic path planning, optimization of neural networks, minimizing
2

weights of Truss Structures, task scheduling over cloudlets, optimization of integrated

process planning, dynamic vehicle routing, susceptibility mapping of forest fire, etc.
[2] Engineers and researchers have been working with the implementation of me-
taheuristic algorithms for a long time. As a result, metaheuristic algorithms such as
Genetic Algorithm(GA), Flower Pollination Algorithm(FPA), Particle Swarm Opti-
mization (PSO), Jaya Algorithm(JA), Hill Climbing(HC), Cuckoo Search (CS), have
become very popular in the field of optimization. Metaheuristic algorithm based op-
timizations are also being mentioned as metaheuristic optimization now [3]. Metaheu-
ristic based optimizations possess stochastic characteristics [4]. These algorithms
have a wide range of applications. Every metaheuristic algorithm has unique charac-
teristics. While performing an optimization, there may be a need for an algorithm with
particular characteristics to solve the optimization problem. So, a specific algorithm
with some characteristics may not be enough to solve a specific optimization problem.

Selecting the most suitable metaheuristic algorithm is vital in optimization. Also,

it's quite a challenging task, especially when the optimization problem is complex [5].
So, it's necessary to have ideas about the comparison of different metaheuristic algo-
rithms. The comparisons of the algorithms help to find out their unknown characteris-
tics. For this reason, a comparison between two of the metaheuristic algorithms have
been done, which are Jaya Algorithm (JA) and Cuckoo Search (CS) Algorithm. There
has been a minimization process of some functions. The coding of the algorithms has
been done in Java using Intellij IDEA Community Edition 2020.2.3. The result of the
comparison will be helpful to determine which one is more suitable for a specific
optimization problem.

The next sections will have a detailed discussion of the algorithms and experi-
ments. In section 2, there will be an overview of JA and CS. Section 3 will discuss the
related works and applications of JA and CS. Section 4 will discuss the methodolo-
gies. Section 5 will describe the tests and results, and related data. Section 6 will ana-
lyze the 2 algorithms based on the test results. Finally, section 7 will conclude the
overall discussions with future works.

2 Background

Jaya Algorithm (JA) and Cuckoo Search (CS) Algorithm, both are efficient in gen-
erating test suite, trying to reach the best solutions and avoid worse solutions.

2.1 Jaya Algorithm

JA is a global algorithm in the field of optimization. It’s one of the recent metaheuris-
tic algorithms. The algorithm is powerful and simple to implement. JA has some ad-
vantages. The potential solution for JA is based on finding the most optimum solution
and avoiding the worst solution. JA doesn't require tuning [6]. It's because, mainly the
common controlling parameters are required for JA. There’s no need for control pa-
rameters that are specific to the algorithm.
 3

2.2 Cuckoo Search Algorithm

CS is an evolutionary optimization algorithm [7]; it’s based on swarm intelligence.
Yang and Deb, being inspired by cuckoos' natural behaviors, developed this algo-
rithm. The algorithm mimics the behavior of some cuckoo species' obligate brood
parasitism where they lay eggs in other host birds' nests [8]. Obligate brood parasitism
is a special characteristics of cuckoos. As a bird, cuckoo’s behavior and activity are
fascinating. It’s not only because of their beautiful sounds, but also due to their ag-
gression in reproduction strategy. Mature cuckoos lay their eggs in the nests of other
host birds. Basically, this algorithm is based on how cuckoos lay eggs and breed. If
the host birds find out that the eggs aren’t theirs, they will either throw away these
unknown eggs or leave their nests and build up new nests somewhere else.

The CS follows 3 important rules:

1. Each of the cuckoos lays an egg at a time and drops the egg in a nest chosen
randomly.
2. The nest having high-quality eggs is the best one. This nest carries over to the
next stage or generation.
3. There's a fixed number of available nests of host birds. The probability Pa Ɛ [0,
1] determines whether the host bird will discover the cuckoo's laid egg or not
[9].

3 Related Works and Applications

This section discusses the existing works of JA and CS. JA and CS have been im-
plemented in various fields and applications.

3.1 Related Works and Applications of Jaya Algorithm

Jaya algorithm (JA) has a wide range of applications. Researchers have implemented
the algorithm for adaptive control for the problem of surge tank indirectly. They have
scrutinized the adaptive controls based on JA for the surge tanks’ nonlinear models
and components having nonlinearities. Usually, the formulations of the controllers are
single objective optimizing problems with different controlling variables to find out
the optimal solutions for satisfying JA’s different constraints. This approach ensures
improvement in randomness and performance of systems [10].

JA is an effective option to extract different PV model parameters [11]. This pro-

cess controls and tracks maximum power point on photovoltaic systems, and also
identifies reliable and accurate model parameters of PV modules and cells. This pro-
cess uses an improved JA that quantifies the individual functioning in the population.
Each individual selects various evolution strategies depending on probability. The
strategies are designed for exploitation abilities and balancing exploration for the
search process. The quantified performance is for constructing the searching direction
by selecting the exemplar. This process ensures an improved population for exploring
4

better solutions; for this, around the present best solution, it introduces a perturbation
mechanism, which is self-adaptive and chaotic.

JA can be implemented for load balancing in the cloud. Cloud computing has vari-
ous challenges like automated resource provisioning, server consolidation, event con-
tent dissemination, security, virtual machine migration, etc. For load balancing in the
cloud, the challenges are decreasing response time, decreasing service request time of
data center, improving the system’s overall performance, etc. JA uses less controlling
parameters and ensures a very good optimized result [12]. This approach has been
proved to have great efficiency while comparing with other approaches.

Researchers have used the JA to perform fuzzy analysis using approaches based on
pressure on benchmark networks [13]. Using fuzzy analysis, we can understand the
uncertainty in different independent parameters of the network of water distribution.
The parameters include pipe roughness values, nodal demands, pipe diameters, reser-
voir heads, etc. Obtaining the dependent parameters’ membership functions are based
on considering uncertain independent parameters’ membership functions. According
to the method of Impact Table, there’s supposed to be a repetitive analysis due to the
monotonous relationship between independent and dependent parameters. Methods
based on optimization are more useful for fuzzy analysis when there’s a non-
monotonous relationship between independent and dependent parameters. JA has
been found to be an efficient option for optimization. The analysis can be done by
using a hydraulic model in EPANET, linked up with MATLAB for optimization.

JA can be implemented for tuning PID controllers for DC servo motor’s position
control [14]. Here, the unit step input’s integral of squared error or ISE is the perfor-
mance index. Using JA, ISE is minimized for obtaining the controller settings. Tuning
based on JA ensures satisfactory response.

3.2 Related Works and Applications of Cuckoo Search Algorithm

Cuckoo Search (CS) based applications have shown very good efficiency in solving
optimization problems. This algorithm provides better solutions as compared to many
other algorithms.

It’s possible to implement enhanced scatter search algorithms using CS [15]. An

example is the problem of traveling salesman using improved and original scatter
search. The improved edition of the scatter search algorithm performs better than the
original one.

CS is an efficient algorithm for solving nurse schedule problems. Lim Huai Tein
used this algorithm for nurse scheduling, which is very useful in healthcare institu-
tions. Also, with the CS, we can solve problems in manufacturing optimization. CS
has been effective in optimizing machine parameters in operations, and has provided
better solutions as compared to other algorithms [16].
 5

Quantum Inspired CS is an improved CS developed by A. Layeb [17]. This method

is based on CS and Principles of Quantum Computing. The process includes defining
the algorithm with a proper representation scheme, allowing the application of the
algorithm on combinatorial optimizing tasks and some principles of quantum compu-
ting, such as measurement, qubit state superposition, interference, representation, etc.
Representing the quantum solutions with a probability, it’s possible to code the solu-
tions. This improved algorithm ensures efficient and optimal solutions with the num-
ber of iterations and population size.

Another application of the CS is a flow of OP-AMP optimization assisted by a pol-

ynomial metamodel of 3 steps [18]. This improved algorithm provides solutions to the
issues of inefficient system performance of optimized OP-AMPs. The CS provides
the desired optimization results, ensuring a design flow for OP-AMP optimization.
The process estimates the performance of OP-AMP; for this, it generates extremely
accurate and ultra-fast polynomiameta models and facilitates quick time-domain sim-
ulating system of a metamacromodel of OP-AMP. These are integrated into a module
of Verilog-AMS.

The particle approach based on CS helps to achieve energy efficiency in multi-

modal objective functions and wireless sensor networks [19]. This approach formu-
lates network optimization. This process randomly deploys the nodes, and organizes
those as static clusters using CS. The collection and aggregation of information are
done after the selection of cluster heads. Then by the generalized algorithm of particle
approach, this process forwards the information to the base station. CS helps to select
cluster heads and form clusters among sensor nodes. This approach provides compa-
rable results as compared to simulation results of LEACH protocols. Sensor net-
work’s longevity increases due to this protocol. Also, complications in chain for-
mation reduces due to this approach.

4 Methodologies

4.1 Implementing Jaya Algorithm

While implementing the JA, the process is based on searching for the best solution
and avoiding the worst solution for a certain problem. In the process, there are basic
parameters, such as the size of the population, termination condition, number of de-
sign variables, etc. The termination condition is usually the maximum number of
iteration. Maximizing or minimizing an objective function f(x) is the primary objec-
tive of this algorithm [20]. Suppose we have m design variables at iᵗ ͪ iteration. Also,
there are n candidate solutions. The best-obtained candidate solution for f(x) from all
candidate solutions is represented as f(x)best. Similarly, we represent the worst candi-
date solution as f(x)worst. We use Xj,k,i to represent the jᵗ ͪ designing variable for kᵗ ͪ
candidate at iteration number i. The modification of Xj,k,i is X'j,k,i. We do this modifi-
cation by the following equation:
6

X'j,k,i = Xj,k,i + r1,j,i (Xj,best,i - |X j,k,i|) – r2,j,i (Xj,worst,i - |X j,k,i|) (1)

In the equation, Xj,best,i represents j variable's value for the best candidate, and
Xj,worst,i represents j variable's value for the worst candidate. r 1,j,i and r2,j,i are two
random numbers ranging from 0 to 1 [21]. With this equation, the search moves to-
wards the best solution. Depending on which one is bigger between X'j,k,i and Xj,k,i,
we update the solution. The process continues as a loop. As the iteration keeps mov-
ing forward, the solution becomes more and more optimum.

4.2 Implementing Cuckoo Search Algorithm

The CS has some stages. Usually, we begin with an objective function f(x). Then
initial population is generated having n host nests xi (i=1,2,….,n). The search contin-
ues till the maximum generation. In each generation, a cuckoo is selected randomly
by lévy flight. We evaluate the fitness or quality of the cuckoo. We choose a nest
randomly. Then we compare the fitnesses of cuckoo and nest and replace the one
having lower fitness [22].

While generating new solution x(t+1) for, suppose cuckoo i, we can use the follow-
ing equation,

x(t+1) = x(t) + α Lévy(β) (2)

We use the product meaning entry-wise walk during multiplications. There's

esentially a random walk due to the Lévy flights [23]. We draw random steps of Lévy
flights from a distribution of Lévy for large steps.

Lévy ~ u = t-1- β (0 < β ≤ 2) (3)

This has an infinite variance, and with it, there's an infinite mean. The consecutive
steps or jumps of a cuckoo develop a random walk process following a power-law
and step-length distribution having a heavy tail. We abandon the worst nest's fraction
Pa to help to build up new nests at new locations by randomly walking and mixing.
The eggs or solutions get mixed by the random permutation depending on the differ-
ence or similarity to the host eggs. The step size sample generation isn’t trivial using
the Lévy flights. A simple scheme can be shown as:

xi(t+1) = xi(t) + α Lévy(β) ~ 0.01 (u / |v|) (x i(t) – xb(t)) (4)

Here, we have drawn u and v from normal distribution.

u = N(0, σu2) (5)

v = N(0, σv2) (6)

 7

1/ β
σu = (7)

σv = 1 (8)

Here, we have used Γ as the standard Gamma function.

5 Tests and Results

All tests of the functions have been done on Intellij IDEA Community Edition
2020.2.3. The tests were performed on Intel(R) Core(TM) i5-8250U (1.60 GHz, 3.4
GHz) with 8 GB of DDR4 RAM on Windows 10 operating system. We have done the
minimization of the functions and updated the minimized values and time perfor-
mances in milliseconds. Each test has been done with 1000 iterations.

For JA, there was no need for any fixed parameter values. After taking the initial
population, the process had moved towards the next iterations, where the population
was updated using equation (1). For CS, there were some fixed parameter values.
Probability, Pa = 0.25. σu = 0.6969 was used based on previous study[24]. The pa-
rameters varied in the iterations.

We have used 11 functions for the tests as shown in table 1.

Table 1. Benchmark functions

No. Functions Conditions

1. f1(x) = x12 – x1x2 + x22 + 2x1 + 4x2 +3 -100 ≤ x1, x2 ≤ 100

2. f2(x) = x12 + 2x22 – 0.3cos(3πx1) – 0.4cos(4πx2) + 0.7 -100 ≤ x1, x2 ≤ 100

3. f3(x) = (x12 + x2 – 11)2 + (x1 + x22 – 7)2 -5 ≤ x1, x2 ≤ 5

4. f4(x) = |x12 + x22 – 2x1x2| + |sinx1| + |cosx2| -500 ≤ x1, x2 ≤ 500

5. f5(x) = 10x4 – 8x2 + 12x + 16 -100 ≤ x ≤ 100

6. f6(x) = 17x5 – 11x2 + 11x + 6 -100 ≤ x ≤ 100

7. f7(x) = 3x12 + 5x22 – 0.6cos(3πx1+4πx2) + 4 -5 ≤ x1,x2 ≤ 5

8. f8(x) = 3x12 + 7x22 + 15(sin2x1+sin2x2) -100 ≤ x1,x2 ≤ 100

9. f9(x) = (2.7 – x1 + x1x2)2 + (1.85 – x1 + x1x22)2 + (3.1 – x1 + x1x23)2 -100 ≤ x1,x2 ≤ 100

10. f10(x) = (x12 + x2 -19)2 + (3x1 + x22 – 16)2 -100 ≤ x1,x2 ≤ 100

11. f11(x) = 104x12 + x22 – (x12 + x22) + {10-4(x12 + x22)} -100 ≤ x1,x2 ≤ 100
8

Table 2. Results for Jaya Algorithm for minimized problems

No. Best Results Worst Mean Time (ms)

f1 -6.33 33.2 -2.934 53.6
f2 0 0 0 59
f3 0 4.3 0.891 56.3
f4 1.00 1.41 1.116 64
f5 5.23 5.23 5.23 54.3
f6 -1.699 x 1011 -1.677 x 1011 -1.692 x 1011 74.9
f7 3.4 3.45 3.407 63.46
f8 0 0.01 0.001 65.55
f9 0.68 2.88 2.265 83.67
f10 0 1.02 0.104 59.1
f11 0 0.84 0.107 44.1

Table 3. Results for Cuckoo Search Algorithm for minimized problems

No. Best Results Worst Mean Time (ms)
f1 -6.33 -6.33 -6.33 33.5
f2 0 0 0 35.1
f3 0 0 0 32.9
f4 1.00 1.15 1.055 151.5
f5 5.23 5.23 5.23 29.6
f6 -1.7 x 1011 -1.7 x 1011 -1.7 x 1011 672391
f7 3.4 3.4 3.4 40.85
f8 0 0 0 37.5
f9 0 0.68 0.612 99.3
f10 0 0 0 40.5
f11 0 0 0 26.5

Table 2 and Table 3 show the test obtained results. Each test was run 20 times and
the best-minimized value, the worst minimized value, mean value, and the execution
time were reported. The execution time of the two algorithms are shown in the tables.
Also, Fig. 1 and Fig. 2 show the comparison of the two algorithms in terms of mean
values and execution times for the functions.
 9

Jaya Cuckoo Search

5.23
5.23

3.407
3.4
1.116
1.055
0.891

0.001
0
0

0
F1 F2 F3 F4 F5 F6 F7 F8
Mean value

-2.934
-6.33

Fig. 1. Performance Comparison between Jaya algorithm and Cuckoo Search in term of test
size

Jaya Cuckoo Search

151.5
Times in milliseconds

99.3
83.67
65.55
63.46

59.1
56.3

64

54.3
53.6

59

40.85

44.1
40.5
37.5
35.1
33.5

32.9

29.6

26.5

F1 F2 F3 F4 F5 F7 F8 F9 F10 F11
Functions

Fig. 2. Execution time comparison between Jaya algorithm and Cuckoo Search.

Fig. 3 shows the screenshot of a Cuckoo Search program done in Intellij IDEA
Community Edition 2020.2.3. The coding of the algorithms has been done in Java.
10

Fig. 3. Screenshot of the part of a program of the Cuckoo Search based function solution with
output in Intellij IDEA Community Edition 2020.2.3

6 Discussions
As observed from the tests, JA has performed well consistently for all the 11 func-
tions. CS hasn’t been consistent like JA, but the mean values determined from the
tests of the CS were nearer to the best-minimized values as compared to JA for many
functions. As shown in table 2 and 3, for 8 of the 11 functions, CS required less exe-
cution time as compared to that of JA but the differences weren’t much. For 3 func-
tions, which are function 4 (f4), function 6 (f6), and function 9 (f9), JA took less execu-
tion time as compared to that of CS. But in these cases, the time differences were
larger as compared to those of other functions. In the case of function 6 (f6), the aver-
age execution time of JA was 74.9 ms, but the average execution time of CS algo-
rithm was 672391 ms. Here, the difference is huge. For function 4 (f4), CS takes
151.5 ms, where JA takes only 64 ms. But when JA takes larger execution time as
compared to CS, the difference isn’t much. There has been no case in the tests where
JA took equal to or more than twice the execution time of CS, according to Fig 2. So,
it indicates that CS may take much larger execution time for some functions as com-
pared to JA. Considering this case, JA is more consistent.
CS showed better performance for 8 of the 11 functions in terms of mean values
and execution times as shown in table 2 and 3, but lacked consistency as compared to
JA. For function 9 (f9), CS was better than JA, but the exection time was a little high-
er. Regarding execution time, in Fig. 2, CS’s performance isn’t always efficient for all
functions. There are some functions, where CS performs at a much slower speed. For
some functions, CS can take extremely large execution time, as shown in the case of
function 6 (f6). Regarding these issues, JA is a simpler and more consistent algorithm,
providing solutions to most functions with an acceptable execution time. But regard-
ing mean values, JA performs well but is less optimum as compared to CS. From the
 11

overall scenario, JA is more consistent due to its performances in acceptable ranges

for most functions.

7 Conclusion and Future Works

This paper evaluated the performances of CS and JA using the benchmark func-
tions in terms of mean values and execution times. Each of the algorithms has shown
impressive results in many cases. There has also been complexities for both algo-
rithms regarding execution time and mean values in some cases. In the overall scenar-
io, JA has turned out to be a more consistent algorithm. The analysis of these 2 algo-
rithms will help engineers to select the right algorithm for specific problems.
As for the future works, we can make a detail analysis explaining for which func-
tions CS works well with acceptable execution times and for which functions, CS
struggles to maintain a faster execution time. Also, there are works to be done with
the functions, for which, CS takes much longer execution time.

ACKNOWLEDGMENTS
This research is funded by Universiti Malaysia Pahang (UMP) under grant: “Priori-
tized T-way Test Suite Generation Strategy Based on Chaotic Flower Pollination
Algorithm”, Grant no: RDU190372. We thank UMP for the contribution and sup-
ports.

References

1. Kelley, T.R.: Optimization, an Important Stage of Engineering Design. The

Technology Teacher 69(5), 18-23 (2010).
2. Mansour, N.: Search Algorithms and Applications. (2011)
3. Yang, X.-S.: Metaheuristic Optimization: Algorithm Analysis and Open Problems.
Paper presented at the International Symposium on Experimental
Algorithms, Berlin. Heidelberg,
4. WK Wong, C.I.M.: A Review on Metaheuristic Algorithms: Recent Trends,
Benchmarking and Applications. Paper presented at the 2019 7th
International Conference on Smart Computing & Communications (ICSCC),
5. Y. A. Alsariera, H.S.A., A. M. Nasser, M. A. Majid and K. Z. Zamli: Comparative
Performance Analysis of Bat Algorithm and Bacterial Foraging Optimization
Algorithm using Standard Benchmark Functions. Paper presented at the
2014 8th. Malaysian Software Engineering Conference (MySEC), Langkawi,
Malaysia,
6. Willa Ariella Syafruddin, M.K., Brahim Benaissa: Does the Jaya Algorithm Really
Need No Parameters? . Paper presented at the 10th International Joint
Conference on Computational Intelligence,
12

7. Pandey, H.M.: Jaya a Novel Optimization Algorithm: What, How and Why? Paper
presented at the 2016 6th International Conference - Cloud System and Big
Data Engineering (Confluence),
8. Swati Sharma, B.B.: Adaptive Control using Jaya Algorithm. In: AIP Conference
Proceedings 2136, 020001 2019
9. Kunjie Yua, B.Q., Caitong Yuea, Shilei Gea, Xu Chenc, Jing Lianga: A
performance-guided JAYA algorithm for parameters identification of
photovoltaic cell and module. Applied Energy, Elsevier 237, 241-257
(2019).
10. Subhadarshini Mohanty, P.K.P., Mitrabinda Ray, Subasish Mohapatra: An
Approach for Load Balancing in Cloud Computing Using JAYA Algorithm.
International Journal of Information Technology and Web Engineering
14(1), 27-41 (2019).
11. Sreethu Subrahmanian, R.G.: Fuzzy node flow analysis of water distribution
networks using Jaya algorithm. Paper presented at the IOP Conference Series
Earth and Environmental Science 491:012010
12. ANKIT K. SAHU, J.K.B., V.P. SINGH, S.P. SINGH: JAYA ALGORITHM
BASED TUNING OF PID CONTROLLER. International Journal of
Industrial Electronics and Electrical Engineering 4(12) (2016).
13. Azizah Mohamad, A.M.Z., Nor Erne Nazira Bazin, Amirmudin Udin: Cuckoo
Search Algorithm for Optimization Problems - A Literature Review. Applied
Mechanics and Materials 421, 502-506 (2013).
14. Iztok Fister Jr.*, D.F., Iztok Fister: A comprehensive review of cuckoo search:
variants and hybrids. International Journal Mathematical Modelling and
Numerical Optimisation 4(4), 387-409 (2013).
15. Azizah Binti Mohamad, A.M.Z., Nor Erne Nazira Bazin: Cuckoo Search
Algorithm for Optimization Problems—A Literature Review and its
Applications. Applied Artificial Intelligence: An International Journal 28(5),
419-448 (2014).
16. Al-Obaidi, A.T.S.: Improved Scatter Search Using Cuckoo Search. International
Journal of Advanced Research in Artificial Intelligence 2(2) (2013).
17. Lim Huai Tein, R.R.: Recent Advancements of Nurse Scheduling Models and A
Potential Path. In: Proceedings of the 6th IMT-GT Conference on
Mathematics, Statistics and its Applications (ICMSA2010), Kedah, Malaysia
2010
18. Yildiz, A.R.: Cuckoo search algorithm for the selection of optimal machine
parameters in milling operations. The International Journal of Advanced
Manufacturing Technology 64, 55-61 (2012).
19. Layeb, A.: A novel quantum-inspired cuckoo search for Knapsack problems.
International Journal of Bio-Inspired Computation 3(5) (2011).
20. G. Zheng, S.P.M.a.E.K.: Metamodel-Assisted Fast and Accurate Optimization of
an OP-AMP for Biomedical Applications. Paper presented at the 2012 IEEE
Computer Society Annual Symposium on VLSI, Amherst, MA, USA,
21. Al-Hakam Ayad Salih, A.H.A., Nada Yousif Hashim: Jaya: An Evolutionary
Optimization Technique for Obtaining the Optimal Dthr Value of Evolving
 13

Clustering Method (ECM). International Journal of Engineering Research

and Technology 11(12), 1901-1912 (2018).
22. Dhivya Manian, M.S.: Energy Efficient Computation of Data Fusion in Wireless
Sensor Networks Using Cuckoo Based Particle Approach (CBPA).
International Journal of Communications, Network and System Sciences
4(4), 249-255 (2011).
23. Hongqing ZHENG, Y.Z.: A Novel Cuckoo Search Optimization Algorithm Base
on Gauss Distribution. Journal of Computational Information Systems 8(10),
4193-4200 (2012).
24. Nasser, A.B., Alsewari, A.R.A., Zamli, K.Z.: Tuning of cuckoo search based
strategy for t-way testing. In: International Conference on Electrical and
Electronic Engineering 2015, vol. 9, p. 8948. Journal of Engineering and
Applied Sciences