This is a post-print version of an article published in Expert Systems with Applications.

The final publication is available at

Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Salp Swarm Optimization: a Critical Review

Mauro Castelli1 , Luca Manzoni2 , Luca Mariot3 , Marco S. Nobile4,5,6 ,

and Andrea Tangherloni7
1 NOVA Information Management School (NOVA IMS), Universidade Nova de Lisboa, Campus

de Campolide, 1070-312, Lisboa, Portugal ,

[email protected]
2 Department of Mathematics and Geosciences, University of Trieste, 34127 Trieste, Italy ,
[email protected]
3 Cyber Security Research Group, Delft University of Technology, 2628 CD Delft, The

Netherlands ,
[email protected]
4 Department of Industrial Engineering & Innovation Sciences, Eindhoven University of

Technology, 5612 AZ, Eindhoven, The Netherlands ,

[email protected]
5 SYSBIO/ISBE.IT Centre of Systems Biology, 20126, Milan, Italy
6 Bicocca Bioinformatics, Biostatistics and Bioimaging Centre (B4), 20854, Vedano al Lambro,

Italy
7 Department of Human and Social Sciences, University of Bergamo, 24129, Bergamo, Italy ,
[email protected]

May 7, 2022

Abstract
In the crowded environment of bio-inspired population-based metaheuristics,
the Salp Swarm Optimization (SSO) algorithm recently appeared and immediately
gained a lot of momentum. Inspired by the peculiar spatial arrangement of salp
colonies, which are displaced in long chains following a leader, this algorithm seems
to provide an interesting optimization performance. However, the original work
was characterized by some conceptual and mathematical flaws, which influenced
all ensuing papers on the subject. In this manuscript, we perform a critical review
of SSO, highlighting all the issues present in the literature and their negative effects
on the optimization process carried out by this algorithm. We also propose a
mathematically correct version of SSO, named Amended Salp Swarm Optimizer
(ASSO) that fixes all the discussed problems. We benchmarked the performance
of ASSO on a set of tailored experiments, showing that it is able to achieve better
results than the original SSO. Finally, we performed an extensive study aimed
at understanding whether SSO and its variants provide advantages compared to
other metaheuristics. The experimental results, where SSO cannot outperform
simple well-known metaheuristics, suggest that the scientific community can safely
abandon SSO.

1
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Keywords Metaheuristics; Global Optimization; Bound Constrained Optimization;

Shift Invariant Functions.

1 Introduction
Most of the real-world problems can be formulated in terms of optimization problems,
that is, they can be solved by finding the values that lead to the maximization, or
minimization, of a target fitness function. However, an analytic solution for such
function is seldom available; moreover, these problems are often high-dimensional
and are characterized by several local optima, where simple optimization algorithms
easily get stuck. In this context, bio-inspired population-based global optimization
metaheuristics often proved their effectiveness in the identification of optimal solutions
(i.e., global optima).
A very large number of metaheuristics have been proposed in the latter years, taking
inspiration among others from: Darwinian processes such as Differential Evolution [53],
Evolution Strategies [31], Genetic Algorithms [32], and all related variants [66]; the
emergent collective behavior of groups of animals, including Particle Swarm Optimiza-
tion [20, 52], Artificial Bee Colony [37], Ant Colony Optimization [18], Bat Swarm
Optimization [74], Firefly Algorithm [76], and all related variants [46]); finally, nat-
ural dynamic phenomena such as Gravitational Search [55] and Artificial Immune
Systems [12].
Recently, a metaheuristics named Salp Swarm Optimization (SSO), which is inspired
by salp colonies, was proposed in [43]. Salps are barrel-shaped animals living in swarms
that are organized in long chains, actively looking for phytoplankton. SSO leverages
the metaphor of this peculiar spatial arrangement by translating it into an optimization
algorithm. Specifically, the individuals composing a population are constrained to follow
the leader salp, while the latter is performing the actual exploration for food sources. In
this metaphor, the leader explores the search space for optimal regions—with respect to
the fitness function—while the followers exploit the area surrounding the leader.
SSO quickly became popular as a real-valued [2], discrete [6, 23], and even multi-
objective optimization algorithm [36]. This popularity apparently stems both from its
performance and peculiar functioning paradigm, making the method to stand out among
a plethora of seemingly novel algorithms [17].
In this paper, we show that the original algorithm is characterized by some conceptual
and implementation flaws, hampering the performance of SSO and all derived algorithms
under some specific circumstances. In particular, we will show that the most remarkable
issue is the fact that SSO, in its original formulation, is not shift-invariant. This raises
a problem if the lower bound of the search space is too far from the origin. In this
paper, we first propose an Amended Salp Swarm Optimizer (ASSO), which represents a
correct reformulation of the SSO algorithm as was probably intended by its creators.
Still, our objective is not to promote the use of ASSO in the scientific community.
Indeed, while ASSO fixes the mathematical flaws of SSO, it is still not grounded on
any theoretical basis. Thus, ASSO would be one of the several metaheuristics based on
some natural metaphor and built by combining search operators that do not rely on any
theoretical property [49]. For this reason, we subsequently assess the performance of
ASSO and SSO against simple and well-known metaheuristics on a set of benchmark
functions. This analysis aims at understanding whether the scientific community could
abandon the use of SSO and ASSO, or if these algorithms provide some advantages
with respect to simpler metaheuristics widely established in this research field. As a side

2
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

note, SSO changed name several times in the literature: indeed, we can find it as Salp
Swarm Algorithm [56], Salp Swarm Optimizer [22, 58], Salp Swarm Optimization [11],
Salp Optimization Algorithm [64], and so forth. Incidentally, in the highlights of the
original paper [43], the authors state that the official name of the algorithm is Salp
Swarm Optimizer, although the acronym SSA is used in the rest of the article. In this
paper, we will use the name Salp Swarm Optimization and the acronym SSO to prevent
confusion. We wish to emphasize that this work aims at discussing the limitations
and issues of the original SSO algorithm, showing how even a simple Random Search
(RS) algorithm can outperform the original incorrect SSO algorithm, under specific
circumstances. Therefore, we think that our results and findings raise an alert on the
existing SSO literature.
This paper is structured as follows.
Section 2 reviews the related works highlighting the flaws and limitations of some
of the recently defined metaheuristics. Section 3 provides a critical review of SSO, in
particular concerning the methodological issues with the updating rule of the position
of the leader salp and the physically-inspired motivation for the updating rule of the
positions of the follower salps. Section 4 presents an experimental evaluation of the
improved version of SSO that fixes some of the raised issues, and compares the result
with the original version on a set of standard benchmark functions. Subsequently,
we analyze the performance of SSO and its variants on the CEC 2017 benchmark
function suite against two commonly used metaheuristics, namely Covariance Matrix
Adaptation Evolution Strategy (CMAES) [30] and Differential Evolution (DE) [60].
Finally, Section 5 summarises the main issues of the original SSO algorithm, and
concludes with a broader remark on the nature of the salp metaphor.

2 Related work
Recent years have witnessed the definition of a significant number of metaheuristics
inspired by some natural phenomenon [24]. The common idea behind the definition
of these metaheuristics is to consider a natural process and, subsequently, to design
the underlying metaheuristic by exploiting the natural metaphor observed. After the
publication of a given metaheuristic, it is also common to see a significant number of
scientific papers that use the new metaheuristic to address complex real-world problems
claiming its superior performance compared to existing metaheuristics. Luckily, a
research strand started to criticize the definition of these metaheuristics, observing
that, in most of the cases, their performance cannot be better than the commonly
used evolutionary strategies [69]. Despite the lack of novelty that characterizes these
metaheuristics (i.e., the change concerns only the underlying natural metaphor), a
fundamental issue is that some of the results achieved through them are not reliable.
In this sense, one of the clearest examples is the paper by Weyland [69], in which
the author demonstrated that Harmony Search (HS) [39] cannot be used to successfully
solve a sudoku, thus contradicting the results obtained by Geem [27]. More precisely,
Weyland first proved that HS is a special case of evolution strategies, thus highlighting
the lack of novelty of the metaheuristic. As a consequence, the performance of HS is
always bounded by the performance that can be obtained by evolution strategies. Finally,
Weyland demonstrated that the results achieved in [27] are flawed both from the theoret-
ical and the practical point of view, concluding that there is no reason for the existence
of HS as a novel metaheuristic. Despite the Weyland’s work clearly demonstrated the
lack of novelty of HS, the algorithm is still largely used nowadays. Further, even though

3
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Weyland proved beyond any doubt that HS cannot perform better than evolutionary
strategies, several papers still claim its presumed superior performance [5].
Thus, it seems that practitioners are still deceived by metaheuristics whose novelty
is based only on the use of some natural metaphors. The truth is that HS, and other
related metaheuristics, are simply using a different terminology with respect to the
classic evolutionary strategies. While the lack of novelty did not prevent metaheuristics
based on natural metaphors to be published in well-renowned scientific venues, some
scientific journals are seriously tackling the problem. For instance, Marco Dorigo,
the editor in chief of Swarm Intelligence, published an editorial note [16] stating that
he observed a new trend consisting in “taking a natural system/process and use it as
a metaphor to generate an algorithm whose components have names taken from the
natural system/process used as metaphor”. Dorigo also highlighted that “this approach
has become so common that there are now hundreds of so-called new algorithms that
are submitted (and unfortunately often also published) to journals and conferences
every year”, and concluded his editorial stating that “it is difficult to understand what is
new and what is the same as the old with just a new name, and whether the proposed
algorithm is just a small incremental improvement of a known algorithm or a radically
new idea”.
A similar analysis on this trend appeared in the work by Cruz et al. [14]. There, the
authors first highlighted the vast number of swarm intelligence algorithms developed
by taking inspiration from the behavior of insects and other animals and phenomena.
Subsequently, they showed that most algorithms present common macro-processes
among them, despite the fact that they are inspired by different metaphors. In other
words, the considered metaheuristics are characterized by common issues and features
which happen at the individual level, promoting very similar emergent phenomena [14].
Thus, it is difficult (if not impossible) to claim that such metaheuristics are really novel.
Focusing on specific metaheuristics, some contributions that analyze the behavior
of a given algorithm started to appear [45, 67, 68]. In [68], Villalón et al. thoroughly
investigated the Intelligent Water Drops (IWD) algorithm [33], a metaheuristic proposed
to address discrete optimization problems. The authors demonstrated that the main steps
of the IWD algorithm are special cases of Ant Colony Optimization (ACO) [18]. Thus,
the performance of IWD cannot be better than the best ACO algorithm. Moreover, the
authors analyzed the metaphor used for the IWD definition, and from their analysis, the
metaphor is based on “unconvincing assumptions of river dynamics and soil erosion
that lack a real scientific rationale”. Finally, they pointed out that the improvements
proposed to the IWD algorithm are based on ideas and concepts already investigated in
the literature many years before in the context of ACO.
Niu et al. [45] analyzed the Grey Wolf Optimization (GWO) algorithm [44], and
demonstrated that, despite its popularity, GWO is flawed. In particular, GWO shows
good performance for optimization problems whose optimal solution is 0, while the
same performance cannot be obtained if the optimal solution is shifted. In particular,
when GWO solves the same optimization function, the farther the function’s optimal
solution is from 0, the worse its performance is. Interestingly, GWO was proposed by
the same author of the SSO algorithm analyzed in this paper, and it presents some of the
SSO’s flaws.
Villalón et al. [67] analyzed the popular Cuckoo Search (CS) algorithm [75], a
metaheuristic introduced in 2009. The authors analyzed CS from a theoretical standpoint,
and showed that it is based on the same concepts as those proposed in the (µ + λ)
evolution strategy proposed in 1981 [57]. Further, the authors evaluated the algorithm
and the metaphor used for its definition based on four criteria (i.e., usefulness, novelty,

4
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

dispensability, and sound motivation), and they concluded that CS does not comply with
any of these criteria. Finally, they pointed out that the original CS algorithm does not
match the publicly available implementation of the algorithm provided by the authors
of the algorithm.
This analysis is quite surprising, given the popularity of CS, and it highlights
the need for a thorough investigation of the existing metaheuristics, with the goal of
understanding which of them should be abandoned by the scientific community. Indeed,
the impressive number of metaheuristics published in the literature makes it difficult to
determine whether they really contribute to the advancement of the field. This problem
was pointed out by Martínez et al. [26]. Specifically, the authors investigated whether
the increasing number of publications is correlated with real progress in the field of
heuristic-based optimization. To answer this research question, the authors compared
five heuristics proposed in some of the most reputed journals in the area, and compared
their performance to the winner of the IEEE Congress on Evolutionary Computation
2005. The results showed that the considered methods could not achieve the result of the
competition winner, which was published several years before. Moreover, a comparison
with the state-of-the-art algorithms is often missing, thus making it impossible to
understand the real advantage provided by a new method.
In the same vein, Piotrowski and Napiorkowski [49] highlighted the risk associated
with the definition of new and increasingly complex optimization metaheuristics and the
introduction of structural bias [38]. In particular, the authors focused on two winners of
the CEC 2016 competition, and they found out that each of them includes a procedure
that introduces a structural bias by attracting the population towards the origin. As a final
message, the authors highlighted that some metaheuristics have to be simplified because
they contain operators that structurally bias their search, while other metaheuristics
should be simplified (or abandoned) as they use unnecessary operators.

3 Critical review of SSO

This section outlines the design issues we found in the original definition and implemen-
tation of the SSO algorithm. More precisely, the following issues are discussed here
and in Section 4:
• the update rule of the main salp does not work correctly when one of the dimen-
sions has a lower bound different from zero (Section 3.1). In the experimental
part, we show that simply shifting a 2D sphere function from the origin makes
SSO perform worse than a simple RS (Section 4);
• the physical motivations for the updating rule of the follower salps are incorrectly
derived from Newton’s laws of motion (Section 3.2);
• there is a clear divergence between the algorithm described in [43] and the
available SSO implementation. This makes it difficult, or even impossible, to
compare results from different papers (Section 3.3);
• finally, we experimentally show that the original SSO algorithm has a bias towards
the origin. Since many of the considered benchmark problems have the optimum
in the origin, the results are biased in favor of SSO (Section 4).
In what follows, we will refer to the equations that were introduced in the original
SSO paper by [43].

5
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

3.1 Leader Salp Updating Rule

We assume here that a chain of N different salps moves within a bounded D-dimensional
search space, aiming at identifying the optimal solution.
The first issue is related to the definition of the updating rule for the position of the
leader salp. In Equation 3.1 of the original paper [43], the update for the leader salp
along the j-th dimension, with j = 1, . . . , D, is given as follows:

F j + c1 ((ul j − lb j )c2 + lb j ) if c3 ≥ 0

x1j =  ,

(1)
F j − c1 ((ul j − lb j )c2 + lb j ) if c3 < 0


where the upper and lower bounds of the search space in the j-th dimension are ul j and
lb j , respectively. The value F j ∈ [lb j , ul j ] is the position of the best solution found so
far in the j-th dimension, that corresponds to the best food source. On the other hand, c1
decreases exponentially with the number of iterations according to the following rule:
4` 2
c1 = 2e−( L ) ,

where ` is the current iteration number while L is the total number of iterations. Finally,
both c2 and c3 are random numbers selected uniformly in the range [0, 1].
Here, we can find the first definitional issue, although it did not propagate in the
original source code. Indeed, c3 ∈ [0, 1] implies that the second case of Equation (1) is
never verified. However, in the source code the threshold for c3 is set to 0.5 instead of 0,
which means that the two cases of the updating rule occur with equal probability. This
kind of oversights is usually not a problem, but several successive papers do not correct
the issue (see, e.g., [4, 8, 21, 34, 35, 56, 65, 72, 73]). Further, if a new implementation
following the specifications of [43] is used, instead of the original one released by the
authors, the results might not be comparable. It is currently unknown how many papers
on SSO employ the same implementation of the original paper.
The main issue with the updating rule is however more significant, and it is still
present in the source code and widespread across all existing SSO variants ( [1, 3, 4, 7, 8,
21, 22, 34, 35, 54, 56, 65, 72, 73]). Let us consider the variation with respect to F j , which
is:

c1 ((ul j − lb j )c2 + lb j ) = c1 c2 (ul j − lb j ) + c1 lb j .

Notice that while c1 c2 (ul j − lb j ) gives a value in [0, ul j − lb j ] (i.e., F j + c1 c2 (ul j − lb j )

might remain inside the search space), the value c1 lb j is also added and it can be
arbitrarily large in absolute value. Although this is not the case in the experiments
performed in the original paper, simply shifting the search space by a suitably large
constant might significantly hamper the search process of SSO. The issue could also
potentially affect research in applied disciplines where SSO has been used (e.g., COVID-
19 related research, as proposed by [4]).
As an example, let us suppose that the lower and the upper bounds are of the
same order of magnitude for all the D dimensions, and in particular that lb j = 10k and
ub j = 10k + 1, with j = 1, . . . , D. In other words, the search space is the hypercube
[10k , 10k + 1] j . By using the updating rule given in Equation (1), the position of the
leader salp is updated as

x1j = F j ± c1 c2 (ul j − lb j ) + c1 lb j
. (2)
= F j ± c1 c2 + c1 10k

6
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Since c1 c2 ≤ 2 (recall that c2 ∈ [0, 1]), the position update in Equation (2) is dominated
by the c1 10k term. Consequently, this update will move the leader salp out of the
admissible bounds for most of the values taken by c1 , forcing its position to be clipped
on the borders of the search space most of the times. This implies that SSO is not
4` 2
invariant with respect to translations of the search space. Indeed, given that c1 = 2e−( L ) ,
the salp remains inside the search space [10k , 10k + 1] j only if c1 10k ≤ 1, namely:
!
−( 4` )2 k 1 −( 4` )2
2e L 10 ≤ 1 ⇒ k ≤ log10 e L .
2

The effect is that, depending on the search space, the majority of the updates of the
leader salp can force it on the boundary of the space (due to clipping), with only the later
iterations (with c1 small enough) resulting in the leaders salp moving without clipping.
4` 2
In particular e−( L ) yields the smallest value when ` = L, i.e., at the last iteration and in
this case we obtain that
!
1
k ≤ log10 e−16 ≈ 6.648 .
2

Therefore, k > 6.648 will result in a search space where the position of the leader salp
will always be forced outside of the search space, or equivalently, the leader salp will
continue to “bounce” on the boundaries of the search space.
Similar pathological examples can be found by tweaking the values of the upper
and lower bounds. This observation reveals that the initial value of c1 must be carefully
chosen with respect to the size and the shift of the search space. In other words, SSO is
also not invariant with respect to rescalings of the search space.
Another different issue is that for any dimension j, the quantity c2 (ul j − lb j ) + lb j
ul +lb
has an expected value of j 2 j . When the search space is centered in zero the expected
value is then zero. As we will see in the experimental part, this gives an unfair advantage
to problems where the search space is symmetric (with respect to 0) and the global
optimum is in 0.

3.2 Physically-inspired motivations

The original paper claimed that the definition of the updating rule for the follower salps
is based on the principles of classical mechanics (Newton’s laws). However, there are
important issues concerning the formulation of this rule, as well as the correct use of
Newton’s laws of motion. The equation for the update of the follower salps in SSO is:
1
xij = at2 + v0 t , (3)
2
where it can be assumed that a = aij and v0 = vij at t = 0, for each dimension j (with
j = 1, . . . , D) and follower salp i (with i = 1, . . . , N). In the original paper, the acceleration
a is calculated as:
vfinal
a= , (4)
v0
which is incorrect, since the average acceleration in a time interval ∆t is:
v(t + ∆t) − v(t)
a= .
∆t

7
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Notice that the incorrect formula also gives a dimensionless quantity, instead of a
length divided by a squared time. Notably, this issue is only sometimes corrected (with
a = vfinal − v0 ) in the SSO literature.
In addition, the formula v f inal = x−x 0
t , with x and x0 being the final and initial
positions, respectively, and t being the time interval, is correct for the average speed,
which is however not necessary in all the derivations of the paper. In fact, for computing
the average acceleration, the instantaneous speed must be used instead. Furthermore,
the aim of the derivation is to obtain the final position of the salp, which we cannot use
to compute the average speed. Since the salps are initially still, the original definition of
SSO [43] explicitly uses v0 = 0 which, when substituted in the Equations 3 and 4, gives
an infinite acceleration that would lead the salps outside any boundary of the search
space.
Regardless of these issues, the authors eventually point out that t, in the equations,
corresponds to the iteration number, so that ∆t = 1 and the time term can be cancelled
out from the equations. They conclude that this step leads to the final formula:
1
xij = (xij + xi−1
j ) . (5)
2
Unfortunately, this equation cannot be derived from the previous ones. In fact, the
position of the next salp in the chain never appears before this point and it is not taken
into account in any of the previous derivations.
A correct way to derive the previous formula would be the following. Assume that
the i-th salp is moving toward the current position of the (i − 1)-th salp (in the j-th
dimension) with starting speed of 0, final speed of 1 unit for time step, and constant
acceleration of (xij (t) − xi−1
j (t)) units for time step squared. Hence, the new position after
one time unit can be computed as follows:
1
xij (t + 1) = (xi−1 (t) − xij (t)) + xij (t)
2 j
1
= (xij (t) + xi−1
j (t)).
2
Notice that this is only one of the possible ways to correctly derive the updating rule for
the follower salps, but it is completely unrelated to the biological metaphor exploited
by the authors. This critique only concerns the physical motivations for the definition
of the updating rules, not the updating rule itself. However, the flawed explanation
presented in the original paper is restated into multiple papers [34, 35, 65, 72], without
any significant correction.

3.3 Implementation Issues

It is worth mentioning that both the MATLAB® and Python implementations1 do not
correctly implement the pseudo-code and, unfortunately, do not follow the explana-
tions provided by the authors in the original paper [43]. As a matter of fact, in both
implementations, the authors updated the first half of the population using Equation 1
and the second half using Equation 5. Considering that the salp chain is composed of
N different individuals, the first N2 − 1 salps perform the exploration process for food
sources attracted by the best food source found so far (updated by Equation 1). The
1 The source code of both implementations is available at the following address: http://www.

alimirjalili.com/SSA.html

8
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

N
2 -thindividual is the leader salp that drags the follower salps, which exploit the area
surrounding the leader (updated by Equation 5). In what follows, we will refer to this
implementation as SSO-code.
We modified the implementation of SSO-code by removing the term c1 lb j from
Equation (1), for each dimension j (with j = 1, . . . , D). In such a way, we are able to
avoid as much as possible the clipping step of the salp positions due to the wrong update,
proposed in the original Equation (i.e., Equation 1), which sends the salps out of the
admissible bounds. We will refer to this version as ASSO.

4 Experimental Evaluation
As a first batch of tests, we compared the performance of a simple Random Search
(RS), SSO, SSO-code, and ASSO using 2-D standard benchmark functions (i.e., Ackley,
Alpine, Rosenbrock, and Sphere). Then, the RS, SSO, SSO-code, and ASSO were
compared against basic versions of CMAES [30] and DE [60]. We used the standard
version of CMAES and DE implemented in the Pymoo library (Multi-objective Opti-
mization in Python) [10], which allows for easily using both single- and multi-objective
algorithms, exploiting the default parameters proposed by the authors of the library.
Specifically, concerning CMAES, the initial standard deviation was set to 0.5 for each
coordinate, and no restart strategy was applied. Regarding DE, according to the classic
DE taxonomy, the DE/rand/1/bin variant was used with a differential weight F = 0.3
and a crossover rate equal to 0.5.

4.1 Standard benchmark functions

To show that both SSO and SSO-code are not shift-invariant, the search spaces of the
tested benchmark functions were shifted by a large constant (i.e., 109 ). To collect
statistically sound results, for each function, we ran the tested techniques 30 times.
For each iteration, we kept track of the fitness value of the best individual, over the
30 repetitions, to calculate the Average Best Fitness (ABF). For a completely fair
comparison among the different techniques, we fixed a budget of 100 iterations using
50 individuals. Note that the implemented RS randomly generates 50 particles, at each
iteration, without taking into account any information of the previous iterations.
Figure 1a clearly shows that both SSO and SSO-code are not shift-invariant. Indeed,
shifting 2-D standard benchmark functions by a large constant hampered the optimiza-
tion abilities of both SSO and SSO-code. Across all the tested functions, even RS
obtained better results compared to SSO and SSO-code. On the contrary, our proposed
algorithm ASSO – where we simply removed the term c1 lb j – was able to outperform
the other techniques. It is worth reminding that ASSO is not a novel algorithm, but an
amended version of SSO in which the mathematical errors have been corrected.
In order to evaluate whether the achieved results were different also from a statistical
point of view, we applied the Mann–Whitney U test with the Bonferroni correction
[19, 42, 70]. Specifically, we applied this statistical test to independently compare
the results obtained by the techniques on each benchmark function. Thus, for each
benchmark function and for each technique, we built a distribution by considering the
fitness value of the best individual at the end of the last iteration over the 30 repetitions.
The boxplots in Figure 1b show the distribution of the best fitness values, achieved at
the end of the executions of the tested techniques. Figure 1b also reports the results
of the statistical tests by using the asterisk convention. These results indicate that,

9
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Shifted Ackley Shifted Alpine Shifted Rosenbrock Shifted Sphere

100 10 1 101 10 2

10 5
10 2 Random 10 2
ABF

ABF

ABF

ABF
100
SSO 10 8
SSO-code 10 3
10 4
10 1 10 11
ASSO 10 4
0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100
Iteration Iteration Iteration Iteration

(a)

Shifted Ackley Shifted Alpine Shifted Rosenbrock Shifted Sphere

**** **** **** ****
**** **** **** ****
* ns **** ns
**** **** **** ****
ns ns *** ns
*** *** **** ****
10 0.3 150 150
8
0.2 100 100
Fitness

Fitness

Fitness

Fitness
6
4
0.1 50 50
2
0 0.0 0 0
Random

Random

Random

Random
SSO

ASSO

SSO

ASSO

SSO

ASSO

SSO

ASSO
SSO-code

SSO-code

SSO-code

SSO-code
(b)

Figure 1: (a) Comparison of the ABF obtained by the analyzed techniques for each
tested function. The ABF was calculated by using the fitness values of the best individual
over the 30 repetitions. To better appreciate the differences among the tested techniques,
the y-axes are on a logarithmic scale. (b)The boxplots show the distribution of the last
fitness value of the best individual over the 30 repetitions. p-value≤ 0.0001 (****);
0.0001 <p-value≤ 0.001 (***); 0.001 <p-value≤ 0.01 (**); 0.01 <p-value≤ 0.05 (*);
p-value> 0.05 (ns).

generally, there is no statistical difference among SSO and SSO-code. Only considering
the Rosenbrock function a strong statistical difference is present between SSO and SSO-
code. The simple RS obtained similar or even better results than SSO and SSO-code,
while there is always a strong statistical difference between the results achieved by
ASSO and those achieved by the other techniques, demonstrating the effectiveness of
our correction.
Concerning the problem described in Section 3.1, we show that, on a search space
symmetric with respect to zero, the SSO algorithm has a bias toward the origin. In
order to show this behavior, we will use a fitness function that returns a random number
with uniform distribution in [0, 1). For a swarm intelligence algorithm, we would
expect a uniform distribution of particles across the entire search space using such a
fitness function. Stated otherwise, using a random fitness function, the salps should not
converge anywhere, and should randomly wander across the search space. However,
we will show that, following the non-amended equations provided in the original SSO
paper [43], the swarm converges in the origin providing an unfair advantage in the case
of optimization problems whose global optimum lies in x = 0 (and possibly lead to
sub-optimal performance in the case of real-world functions).
Figure 2 shows the result of this test, performed on both SSO and SSO-code with

10
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

 662ZLWKIRRGDWWUDFWRU  662ZLWKRXWIRRGDWWUDFWRU

 

 
6WDUW 6WDUW

 

 
x2

x2
 

 

 

 
                 
x1 x1
 662FRGHZLWKIRRGDWWUDFWRU  662FRGHZLWKRXWIRRGDWWUDFWRU
6HTXHQFHOHDGHU
,WHUDWLRQ
 6WDUW  6WDUW ,WHUDWLRQ
,WHUDWLRQ
,WHUDWLRQ
  ,WHUDWLRQ
,WHUDWLRQ
,WHUDWLRQ
  ,WHUDWLRQ
,WHUDWLRQ
,WHUDWLRQ
 
x2

x2

 

 

 

 
                 
x1 x1

Figure 2: Dynamic behavior of the salp swarm: the leader leaves the feasible region, is
relocated on the boundary, and then the swarm is attracted towards the origin of axis.

and without the food attractor. The Figure reports the positions of all salps during the
first 10 iterations of the optimization; the position of the leader salp is highlighted by an
orange circle, and the initial position is denoted by the text “Start”. According to our
results, the leader salp get attracted toward zero. The same happens for SSO-code: the
leader salp it is inevitably attracted towards the center of the search space. The attraction
towards the origin is even more evident in the case of SSO and SSO-code without food
attraction: the swarm perfectly converges to the origin and no longer moves.

4.2 CEC 2017 benchmark function suite

During the last years, different benchmark function suites have been proposed to test
and compare existing and novel global optimization techniques [63]. The benchmark
functions they contain try to mimic the behavior of real-world problems, which often
show complex features that basic optimization algorithms might not be able to grasp
[25, 63]. Regarding the real-parameter numerical optimization, the IEEE Congress
on Evolutionary Computation (CEC) and the Genetic and Evolutionary Computation
Conference (GECCO) include competitions where complex benchmark function suites
have been designed to test and compare global optimization techniques [13, 15, 28, 29].
Here, we evaluated the performance of the considered optimization techniques
using the CEC 17 benchmark problems for single-objective real-parameter numerical
optimization [9], which were previously used to compare the performance of different

11
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Table 1: Definitions and optimum values of the CEC 2017 benchmark functions designed
for the competition on real-parameter single objective numerical optimization.

Typology No. Function name Opt.

1 Shifted and Rotated Bent Cigar 100
Unimodal
2 Shifted and Rotated Sum of Different Power 200
Functions
3 Shifted and Rotated Zakharov 300
4 Shifted and Rotated Rosenbrock 400
5 Shifted and Rotated Rastrigin 500
Simple 6 Shifted and Rotated Expanded Schaffer F6 600
Multimodal 7 Shifted and Rotated Lunacek Bi-Rastrigin 700
Functions 8 Shifted and Rotated Non-Continuous Rastrigin 800
9 Shifted and Rotated Levy 900
10 Shifted and Rotated Schwefel 1000
11 Zakharov; Rosenbrock; Rastrigin 1100
12 High-conditioned Elliptic; Modified Schwefel; Bent Cigar 1200
13 Bent Cigar; Rosenbrock; Lunacek bi-Rastrigin 1300
14 High-conditioned Elliptic; Ackley; Schaffer F7; Rastrigin 1400
Hybrid 15 Bent Cigar; HGBat; Rastrigin; Rosenbrock 1500
Functions 16 Expanded Schaffer F6; HGBat; Rosenbrock; Modified Schwefel 1600
17 Katsuura; Ackley; Expanded Griewank plus Rosenbrock; Schwefel; Rastrigin 1700
18 High-conditioned Elliptic; Ackley; Rastrigin; HGBat; Discus 1800
19 Bent Cigar; Rastrigin; Griewank plus Rosenbrock; Weierstrass; Expanded Schaf- 1900
fer F6
20 HappyCat; Katsuura; Ackley; Rastrigin; Modified Schwefel; Schaffer F7 2000
21 Rosenbrock; High-conditioned Elliptic; Rastrigin 2100
22 Rastrigin; Griewank; Modified Schwefel 2200
23 Rosenbrock; Ackley; Modified Schwefel; Rastrigin 2300
24 Ackley; High-conditioned Elliptic; Griewank; Rastrigin 2400
Composition 25 Rastrigin; HappyCat; Ackley; Discus; Rosenbrock 2500
Functions 26 Expanded Schaffer F6; Modified Schwefel; Griewank; Rosenbrock; Rastrigin 2600
27 HGBat; Rastrigin; Modified Schwefel; Bent Cigar; High-conditioned Elliptic; Ex- 2700
panded Schaffer F6
28 Ackley; Griewank; Discus; Rosenbrock; HappyCat; Expanded Schaffer F6 2800
29 f15 ; f16 ; f17 2900
30 f15 ; f18 ; f19 3000

metaheuristics [47, 62, 63]. Table 1 reports the tested benchmark problems, which
are based on shifted, rotated, non-separable, highly ill-conditioned, and complex opti-
mization benchmark functions [9]. We optimized each function fk (with k = 1, . . . , 30)
considering the dimensions D = {10, 30}, search space boundaries [−100, 100]D , and a
maximum number of function evaluations MaxFES = D · 104 . For each technique, we
executed 30 independent runs to collect statistically sound results.
Figure 3 depicts the Average Best Fitness (ABF) obtained by the analyzed techniques
for each function fk (with k = 1, . . . , 30) and D = 10, showing that DE was able to obtain
better results than all the SSO-based strategies, including ASSO, and the simple RS
tested here. As one can see, DE outperformed the SSO-based strategies in 27 out of 30
functions. CMAES also obtained better results than the SSO-based strategies in more
than half of the tested functions. As we did for the standard benchmark functions, we
evaluated if the achieved results were also different from a statistical point of view by
using the Mann–Whitney U test with the Bonferroni correction [19, 42, 70]. Thus, we
independently compared the results obtained by the six techniques on each benchmark
function fk (with k = 1, . . . , 30) considering the distributions built using the value of the
best individual at the end of the last iteration over the 30 repetitions. The heatmaps
showed in Figure 4 clearly point out that there is almost always a strong statistical
difference between the results achieved by all the SSO-based strategies, including
ASSO, and those obtained by DE. Specifically, SSO-based strategies were able to obtain
comparable or better results than those reached by DE only for the functions f21 , f25 ,
and f28 . Comparing the results achieved by CMAES and those obtained by SSO-based

12
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Function f1 Function f2 106

Function f3 Function f4 Function f5
3 × 103 Random ASSO
109 1013 2 × 103 SSO CMAES
105 SSO-code DE
107 1010 6 × 102
ABF

104 103
105 107
103 6 × 102
103 104
4 × 102 5 × 102
7 × 102 Function f6 Function f7 Function f8 6 × 103 Function f9 Function f10
4 × 103 3 × 103
103 9 × 102 3 × 103
ABF

9 × 102 2 × 103 2 × 103

8 × 102
6 × 102 103
8 × 102
Function f11 Function f12 Function f13 107 Function f14 Function f15
108 107
108 106 106
104
106 105
ABF

106 105

104 104 104

104
103 103 103
Function f16 Function f17 Function f18 108 Function f19 Function f20
2.8 × 103 2.4 × 103 2.5 × 103
2.6 × 103 2.3 × 103 108 107
2.4 × 103 2.2 × 103 2.4 × 103
2.2 × 103 2.1 × 103 106 2.3 × 103
ABF

2 × 103 106
2 × 103 105 2.2 × 103
1.9 × 103
1.8 × 103 1.8 × 103 104 2.1 × 103
104
1.6 × 103 1.7 × 103 2 × 103

2.45 × 103
Function f21 Function f22 2.85 × 103 Function f23 3.05 × 103 Function f24 Function f25
4 × 103 3 × 103
2.4 × 103 2.8 × 103 2.95 × 103
2.75 × 103 2.9 × 103 4 × 103
2.35 × 103
ABF

3 × 103 2.85 × 103

2.3 × 103 2.7 × 103 2.8 × 103
2.25 × 103 2.65 × 103 2.75 × 103
2.7 × 103 3 × 103
2.6 × 103
5 × 103
Function f26 Function f27 Function f28 4 × 103 Function f29 Function f30
3.4 × 103 4.2 × 103 108
3.35 × 103 4 × 103 3.8 × 103
107
4 × 103 3.3 × 103 3.8 × 103
3.6 × 103
106
ABF

3.25 × 103 3.6 × 103

3.2 × 103 3.4 × 103 105
3.4 × 103
3.15 × 103
3 × 103 3.1 × 103 3.2 × 103 3.2 × 103 104
0 500 1000 1500 2000 0 500 1000 1500 2000 0 500 1000 1500 2000 0 500 1000 1500 2000 0 500 1000 1500 2000
Iteration Iteration Iteration Iteration Iteration

Figure 3: Comparison of the Average Best Fitness (ABF) obtained by the considered
techniques for each function fk (with k = 1, . . . , 30) and D = 10. The ABF was calculated
by using the fitness values of the best individual over the 30 repetitions. Note that the
y-axes are on a logarithmic scale.

strategies, there is a strong statistical difference in more than half of the tested functions.
These results are coherent with the probabilistic re-formulation of the no-free lunch
theorem (NFL) [40], an extension of the original version of the NFL theorem [41, 71],
which prove the validity of the theorem in continuous domains. Thus, according to
the NFL theorem, no algorithm outperforms all the competitors in any optimization
problem.
Increasing the dimensions of the tested benchmark functions from 10 to 30 does not
allow the SSO-based approaches to obtain better results than DE. As a matter of fact,
Figures 5 and 6 clearly show that DE outperformed all the SSO-based approaches in 22
out of 30 benchmarks functions. In the remaining functions (i.e., f4 , f7 , f14 , f18 , f25 ,
f26 , f27 , and f28 ) some SSO-based approaches (generally ASSO) obtained comparable
results than those achieved by DE. However, there is no benchmark function where
all the SSO-based methods were able to obtain similar results to those reached by DE,
confirming the poor performance of all the tested variants of SSO.

5 Conclusion
This paper shows systematic and deep issues in the definition of a widely-cited op-
timization algorithm, namely Salp Swarm Optimization. In particular, the multiple
issues concerning the updating rules, the physically-inspired motivations, and the in-

13
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Function f1 Function f2 Function f3 Function f4 Function f5

Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f6 Function f7 Function f8 Function f9 Function f10
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f11 Function f12 Function f13 Function f14 Function f15
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f16 Function f17 Function f18 Function f19 Function f20
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f21 Function f22 Function f23 Function f24 Function f25
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f26 Function f27 Function f28 Function f29 Function f30
Random 1.0
Confidence interval

SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Random
SSO
SSO-code
ASSO
CMAES
DE
Random
SSO
SSO-code
ASSO
CMAES
DE

Random
SSO
SSO-code
ASSO
CMAES
DE

Random
SSO
SSO-code
ASSO
CMAES
DE

Random
SSO
SSO-code
ASSO
CMAES
DE

Figure 4: Heatmaps showing the p-values of the Mann–Whitney U test with the Bonfer-
roni correction for each function fk (with k = 1, . . . , 30) and and D = 10. The confidence
interval was divided into 4 levels indicating a strong statistical difference, statistical
difference, weak statistical difference, and no statistical difference, respectively.

consistency between the code and the description in the original paper that proposed
Salp Swarm Optimization raise concerns about all ensuing related literature, since the
erroneous derivations and rules are present in most of the published papers on the topic.
Furthermore, it is currently problematic to discern which results can be trusted, which
ones are based on an incorrect implementation, and which papers have an incorrect
description but are using a correct implementation.
The most serious issue analyzed in this paper is perhaps the presence of the lower
bound lb j in the updating rule of the leader salp (see Equation 1). In particular, this
term makes the algorithm not shift-invariant, introducing a severe search bias that

14
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Function f1 Function f2 Function f3 Function f4 Function f5

1011 Random ASSO
1049
108 103 SSO CMAES
109 1039 SSO-code DE
104 9 × 102
107 1029 106 8 × 102
ABF

105 1019 7 × 102

104 103
103 109 6 × 102

Function f6 Function f7 Function f8 Function f9 Function f10

1.4 × 103 104
3 × 103 1.3 × 103
7 × 102 1.2 × 103
2 × 103 104
1.1 × 103 6 × 103
ABF

103
103 9 × 102 4 × 103
6 × 102 103
Function f11 Function f12 Function f13 108
Function f14 Function f15
105 1010
109
109
108
106 107
107
ABF

104 106
105 104 105
104
103
Function f16 Function f17 Function f18 Function f19 Function f20
109
3.75 × 103
109 3.5 × 103
6 × 103 107 3.25 × 103
104 107 3 × 103
ABF

4 × 103 2.75 × 103

3 × 103 105 105 2.5 × 103
2.25 × 103
2 × 103
Function f21 Function f22 Function f23 Function f24 2 × 104 Function f25
2.9 × 103 4 × 103
104
2.8 × 103 4 × 103
2.7 × 103 104
6 × 103
ABF

2.6 × 103
6 × 103
2.5 × 103 4 × 103 3 × 103
2.4 × 103 3 × 103 4 × 103
3 × 103
3 × 103
Function f26 Function f27 Function f28 Function f29 Function f30
109
5 × 103 104
104
107
ABF

4 × 103 6 × 103 104

6 × 103
4 × 103 105
4 × 103
3 × 103
0 2000 4000 6000 0 2000 4000 6000 0 2000 4000 6000 0 2000 4000 6000 0 2000 4000 6000
Iteration Iteration Iteration Iteration Iteration

Figure 5: Comparison of the ABF obtained by the analyzed techniques for each function
fk (with k = 1, . . . , 30) and D = 30 The ABF was calculated by using the fitness values
of the best individual over the 30 repetitions. Note that the y-axes are on a logarithmic
scale.

depends on the distance between the lower bound and the origin. As shown in our
experiments, under some specific circumstances, this factor can significantly affect the
search capabilities of SSO, which was outperformed by a simple random search. If
the term lb j was inadvertently introduced, all the current literature on SSO contains
results not reflecting the intended definition of the algorithm. On the other hand, if
the authors meant to insert the term, the SSO algorithm cannot work in spaces that
have lower bounds too far from 0 in all dimensions. We also compared the described
SSO-based approaches against DE and CMAES for the optimization of the CEC 2017
benchmark functions [9], showing that all the SSO-based versions were outperformed
by a simple DE version on almost all functions. These results highlight, once more,
that SSO and similar algorithms do not give any particular advantages with respect
to widespread and common metaheuristics. Considering all the results discussed in
this work, we expect that more sophisticated metaheuristics, such as Linear population
size reduction Successful History-based Adaptive DE (L-SHADE) [48, 50, 51, 61] that
already showed their superior performance compared to DE, should outperform all the
SSO-based approaches. Thus, based on the evidences of this work, we discourage the
use of SSO by the scientific community. In particular, there is no theory supporting
the convergence properties of SSO and its (supposed) superiority with respect to the
existing metaheuristics. On the contrary, SSO is defined and implemented based on a
wrong mathematical formulation, as discussed in the first part of this paper.
We conclude this paper with a general remark about the metaphor-based approach

15
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

Function f1 Function f2 Function f3 Function f4 Function f5

Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f6 Function f7 Function f8 Function f9 Function f10
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f11 Function f12 Function f13 Function f14 Function f15
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f16 Function f17 Function f18 Function f19 Function f20
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f21 Function f22 Function f23 Function f24 Function f25
Random 1.0

Confidence interval
SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Function f26 Function f27 Function f28 Function f29 Function f30
Random 1.0
Confidence interval

SSO 0.1
SSO-code
0.05
ASSO
CMAES 0.01
DE 0.0
Random
SSO
SSO-code
ASSO
CMAES
DE
Random
SSO
SSO-code
ASSO
CMAES
DE

Random
SSO
SSO-code
ASSO
CMAES
DE

Random
SSO
SSO-code
ASSO
CMAES
DE

Random
SSO
SSO-code
ASSO
CMAES
DE

Figure 6: Heatmaps showing the p-values of the Mann–Whitney U test with the Bonfer-
roni correction for each function fk (with k = 1, . . . , 30) and and D = 30. The confidence
interval was divided into 4 levels indicating a strong statistical difference, statistical
difference, weak statistical difference, and no statistical difference, respectively.

for metaheuristics. As mentioned in the Introduction, a lot of metaheuristic optimization

algorithms have been proposed in the last years, most of them based on a particular
natural process or animal behavior as a metaphor for the exploration and exploitation
phases of the search space. As noted by [59], one of the likely causes of this phenomenon
is the excessive focus on the novelty of such methods in part of the research community
on metaheuristics. This research approach, however, has the downside of shadowing the
true search components of an optimization algorithm with terms and concepts borrowed
from the considered metaphor. Therefore, it can happen that a “novel” metaheuristic
optimization technique turns out to be just another well-known algorithm under a

16
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

heavy disguise. This is the case, for example, of three other recent Swarm Intelligence
algorithms, namely: the Grey Wolf Optimizer, the Firefly Optimization Algorithm, and
the Bat Algorithm, which were shown by [17] to have strong similarities with Particle
Swarm Optimization. While in this manuscript we showed and fixed the methodological
and implementation flaws of SSO, we believe that a closer inspection of the algorithm’s
underlying metaphor would also highlight a strong resemblance to other established
swarm algorithms.

Availability
All source code used for the tests is available on GitLab at the following address:
https://gitlab.com/andrea-tango/asso.

Acknowledgment
This work was supported by national funds through the FCT (Fundação para a Ciência
e a Tecnologia) by the projects GADgET (DSAIPA/DS/0022/2018) and the financial
support from the Slovenian Research Agency (research core funding no. P5-0410).

References
[1] R. Abbassi, A. Abbassi, A. A. Heidari, and S. Mirjalili. An efficient salp swarm-
inspired algorithm for parameters identification of photovoltaic cell models. Energy
Convers. Manag., 179:362–372, 2019.
[2] L. Abualigah, M. Shehab, M. Alshinwan, and H. Alabool. Salp swarm algorithm:
a comprehensive survey. Neural. Comput. Appl., 32:11195—-11215, 2019.
[3] S. Ahmed, M. Mafarja, H. Faris, and I. Aljarah. Feature selection using salp
swarm algorithm with chaos. In Proc. 2nd International Conference on Intelligent
Systems, Metaheuristics & Swarm Intelligence (ISMSI), pages 65–69, 2018.
[4] M. A. Al-Qaness, A. A. Ewees, H. Fan, and M. Abd El Aziz. Optimization method
for forecasting confirmed cases of COVID-19 in China. J. Clin. Med., 9(3):674,
2020.
[5] A. Ala’a, A. A. Alsewari, H. S. Alamri, and K. Z. Zamli. Comprehensive review
of the development of the harmony search algorithm and its applications. IEEE
Access, 7:14233–14245, 2019.
[6] I. Aljarah, M. Mafarja, A. A. Heidari, H. Faris, Y. Zhang, and S. Mirjalili. Asyn-
chronous accelerating multi-leader salp chains for feature selection. Appl. Soft
Comput., 71:964–979, 2018.
[7] I. Aljarah, M. Mafarja, A. A. Heidari, H. Faris, Y. Zhang, and S. Mirjalili. Asyn-
chronous accelerating multi-leader salp chains for feature selection. Appl. Soft
Comput., 71:964–979, 2018.
[8] S. S. Alresheedi, S. Lu, M. Abd Elaziz, and A. A. Ewees. Improved multiobjective
salp swarm optimization for virtual machine placement in cloud computing. Hum.-
centric Comput. Inf. Sci., 9(15), 2019.

17
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

[9] N. H. Awad, M. Z. Ali, J. J. Liang, B. Y. Qu, and P. N. Suganthan. Problem

definitions and evaluation criteria for the CEC 2017 special session and competition
on single objective real-parameter numerical optimization. Technical report,
Nanyang Technological University, Singapore, and Jordan University of Science
and Technology, Jordan, and Zhengzhou University, China, 2016.
[10] J. Blank and K. Deb. pymoo: Multi-objective optimization in python. IEEE
Access, 8:89497–89509, 2020.
[11] Z. Cheng and J. Wang. A new combined model based on multi-objective salp
swarm optimization for wind speed forecasting. Appl. Soft Comput., 92:106294,
2020.
[12] C. A. Coello Coello and N. C. Cortés. Solving multiobjective optimization prob-
lems using an artificial immune system. Genet. Program. Evol. M., 6(2):163–190,
2005.
[13] Congress on Evolutionary Computation (CEC). Special Session & Competitions
on Real-Parameter Single Objective Optimization. http://www.ntu.edu.sg/
home/EPNSugan/index_files/CEC2017/CEC2017.htm, 2017. Online; Ac-
cessed on November 17, 2018.
[14] D. P. F. Cruz, R. D. Maia, and L. N. De Castro. A critical discussion into the core
of swarm intelligence algorithms. Evolutionary Intelligence, 12(2):189–200, 2019.
[15] C. Doerr, F. Ye, S. van Rijn, H. Wang, and T. Bäck. Towards a theory-guided
benchmarking suite for discrete black-box optimization heuristics: profiling (1+
λ) ea variants on onemax and leadingones. In Proc. Genetic and Evolutionary
Computation Conference (GECCO), pages 951–958, New York, NY, USA, 2018.
ACM.
[16] M. Dorigo. Swarm intelligence: A few things you need to know if you want to
publish in this journal, 2016.
[17] M. Dorigo. Grey wolf, firefly and bat algorithms: Three widespread algorithms
that do not contain any novelty. In Proc. 12th International Conference on Swarm
Intelligence (ANTS), Lecture Notes in Computer Science, pages 121–133. Springer
Nature, 2020.
[18] M. Dorigo, M. Birattari, and T. Stutzle. Ant colony optimization. IEEE Comput.
Intell. M., 1(4):28–39, 2006.
[19] O. J. Dunn. Multiple comparisons among means. J. Am. Stat. Assoc., 56(293):52–
64, 1961.
[20] R. Eberhart and J. Kennedy. A new optimizer using particle swarm theory. In
Proc. 6th International Symposium on Micro Machine and Human Science (MHS),
pages 39–43. IEEE, 1995.
[21] S. Ekinci and B. Hekimoglu. Parameter optimization of power system stabilizer
via salp swarm algorithm. In Proc. 5th International Conference on Electrical and
Electronic Engineering (ICEEE), pages 143–147. IEEE, 2018.
[22] A. A. El-Fergany. Extracting optimal parameters of PEM fuel cells using salp
swarm optimizer. Renew. Energy, 119:641–648, 2018.

18
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

[23] H. Faris, M. M. Mafarja, A. A. Heidari, I. Aljarah, A. M. Al-Zoubi, S. Mirjalili,

and H. Fujita. An efficient binary salp swarm algorithm with crossover scheme for
feature selection problems. Knowl-Based Syst., 154:43–67, 2018.

[24] S. Fong. Meta-zoo-heuristic algorithms. In 2017 Seventh International Conference

on Innovative Computing Technology (INTECH), pages 3–8. IEEE, 2017.
[25] M. Gallagher. Towards improved benchmarking of black-box optimization algo-
rithms using clustering problems. Soft Comput., 20(10):3835–3849, 2016.

[26] C. García-Martínez, P. D. Gutiérrez, D. Molina, M. Lozano, and F. Herrera.

Since cec 2005 competition on real-parameter optimisation: a decade of research,
progress and comparative analysis’s weakness. Soft Computing, 21(19):5573–5583,
2017.
[27] Z. W. Geem. Harmony search algorithm for solving sudoku. In International
Conference on Knowledge-Based and Intelligent Information and Engineering
Systems, pages 371–378. Springer, 2007.
[28] Genetic and Evolutionary Computation Conference (GECCO). GECCO Work-
shop on Real-Parameter Black-Box Optimization Benchmarking (BBOB). http:
//numbbo.github.io/workshops/BBOB-2018/, 2018. Online; Accessed on
November 17, 2018.

[29] N. Hansen, A. Auger, O. Mersmann, T. Tusar, and D. Brockhoff. COCO: A

platform for comparing continuous optimizers in a black-box setting. arXiv
preprint arXiv:1603.08785, 2016.
[30] N. Hansen and A. Ostermeier. Adapting arbitrary normal mutation distributions
in evolution strategies: The covariance matrix adaptation. In Proc. International
Conference on Evolutionary Computation (CEC), pages 312–317. IEEE, 1996.
[31] N. Hansen and A. Ostermeier. Completely derandomized self-adaptation in evolu-
tion strategies. Evol. Comput., 9(2):159–195, 2001.
[32] J. H. Holland. Adaptation in Natural and Artificial Systems: An Introductory
Analysis with Applications to Biology, Control and Artificial Intelligence. MIT
Press, Cambridge, MA, USA, 1992.
[33] H. S. Hosseini. Problem solving by intelligent water drops. In 2007 IEEE congress
on evolutionary computation, pages 3226–3231. IEEE, 2007.
[34] A. G. Hussien, A. E. Hassanien, and E. H. Houssein. Swarming behaviour of salps
algorithm for predicting chemical compound activities. In Proc. 8th International
Conference on Intelligent Computing and Information Systems (ICICIS), pages
315–320. IEEE, 2017.
[35] R. A. Ibrahim, A. A. Ewees, D. Oliva, M. Abd Elaziz, and S. Lu. Improved salp
swarm algorithm based on particle swarm optimization for feature selection. J.
Ambient Intell. Humaniz. Comput., 10(8):3155–3169, 2019.
[36] V. Kansal and J. S. Dhillon. Emended salp swarm algorithm for multiobjective
electric power dispatch problem. Appl. Soft Comput., 90:106172, 2020.

19
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

[37] D. Karaboga, B. Gorkemli, C. Ozturk, and N. Karaboga. A comprehensive survey:

artificial bee colony (ABC) algorithm and applications. Artif. Intell. Rev., 42(1):21–
57, 2014.

[38] A. V. Kononova, D. W. Corne, P. De Wilde, V. Shneer, and F. Caraffini. Structural

bias in population-based algorithms. Information Sciences, 298:468–490, 2015.
[39] K. S. Lee and Z. W. Geem. A new meta-heuristic algorithm for continuous
engineering optimization: harmony search theory and practice. Computer methods
in applied mechanics and engineering, 194(36-38):3902–3933, 2005.

[40] A. J. Lockett and R. Miikkulainen. A probabilistic reformulation of no free lunch:

Continuous lunches are not free. Evol. Comput., 25(3):503–528, 2017.
[41] W. G. Macready and D. H. Wolpert. What makes an optimization problem hard?
Complexity, 1(5):40–46, 1996.

[42] H. B. Mann and D. R. Whitney. On a test of whether one of two random variables
is stochastically larger than the other. Ann. of Math. Stat., pages 50–60, 1947.
[43] S. Mirjalili, A. H. Gandomi, S. Z. Mirjalili, S. Saremi, H. Faris, and S. M. Mirjalili.
Salp swarm algorithm: A bio-inspired optimizer for engineering design problems.
Adv. Eng. Softw., 114:163–191, 2017.

[44] S. Mirjalili, S. M. Mirjalili, and A. Lewis. Grey wolf optimizer. Advances in

engineering software, 69:46–61, 2014.
[45] P. Niu, S. Niu, L. Chang, et al. The defect of the grey wolf optimization algorithm
and its verification method. Knowledge-Based Systems, 171:37–43, 2019.

[46] M. S. Nobile, P. Cazzaniga, D. Besozzi, R. Colombo, G. Mauri, and G. Pasi. Fuzzy

Self-Tuning PSO: A settings-free algorithm for global optimization. Swarm Evol.
Comput., 39:70–85, 2018.
[47] M. S. Nobile, A. Tangherloni, L. Rundo, S. Spolaor, D. Besozzi, G. Mauri, and
P. Cazzaniga. Computational intelligence for parameter estimation of biochemical
systems. In Proc. Congress on Evolutionary Computation (CEC), pages 1–8. IEEE,
2018.
[48] A. P. Piotrowski. L-SHADE optimization algorithms with population-wide inertia.
Inf. Sci., 468:117–141, 2018.
[49] A. P. Piotrowski and J. J. Napiorkowski. Some metaheuristics should be simplified.
Information Sciences, 427:32–62, 2018.
[50] A. P. Piotrowski and J. J. Napiorkowski. Step-by-step improvement of JADE and
SHADE-based algorithms: Success or failure? Swarm Evol. Comput., 43:88–108,
2018.
[51] R. Polakova. L-SHADE with competing strategies applied to constrained opti-
mization. In proc. IEEE congress on evolutionary computation (CEC), pages
1683–1689. IEEE, 2017.
[52] R. Poli, J. Kennedy, and T. Blackwell. Particle swarm optimization. Swarm Intell.,
1(1):33–57, 2007.

20
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

[53] K. Price, R. M. Storn, and J. A. Lampinen. Differential evolution: a practical

approach to global optimization. Springer Science & Business Media, 2006.
[54] M. H. Qais, H. M. Hasanien, and S. Alghuwainem. Enhanced salp swarm algo-
rithm: Application to variable speed wind generators. Eng. Appl. Artif. Intell.,
80:82–96, 2019.
[55] E. Rashedi, H. Nezamabadi-Pour, and S. Saryazdi. GSA: a gravitational search
algorithm. Inf. Sci., 179(13):2232–2248, 2009.

[56] G. I. Sayed, G. Khoriba, and M. H. Haggag. A novel chaotic salp swarm algorithm
for global optimization and feature selection. Appl. Intell., 48(10):3462–3481,
2018.
[57] H.-P. Schwefel. Numerical optimization of computer models. John Wiley & Sons,
Inc., 1981.

[58] A. M. Shaheen and R. A. El-Sehiemy. A multiobjective salp optimization algorithm

for techno-economic-based performance enhancement of distribution networks.
IEEE Syst. J., 15(1):1458–1466, 2020.
[59] K. Sörensen. Metaheuristics - the metaphor exposed. Int. Trans. Oper. Res.,
22(1):3–18, 2015.

[60] R. Storn and K. Price. Differential evolution–a simple and efficient heuristic for
global optimization over continuous spaces. J. Glob. Optim., 11(4):341–359, 1997.
[61] R. Tanabe and A. S. Fukunaga. Improving the search performance of SHADE
using linear population size reduction. In prc. IEEE congress on evolutionary
computation (CEC), pages 1658–1665. IEEE, 2014.

[62] A. Tangherloni, L. Rundo, and M. S. Nobile. Proactive particles in swarm optimiza-

tion: A settings-free algorithm for real-parameter single objective optimization
problems. In Proc. Congress on Evolutionary Computation (CEC), pages 1940–
1947. IEEE, 2017.

[63] A. Tangherloni, S. Spolaor, P. Cazzaniga, D. Besozzi, L. Rundo, G. Mauri, and

M. S. Nobile. Biochemical parameter estimation vs. benchmark functions: a
comparative study of optimization performance and representation design. Appl.
Soft Comput., 81:105494, 2019.
[64] F. Tian, H. Wei, X. Li, M. Lv, and P. Wang. An improved salp optimization
algorithm inspired by quantum computing. In Proc. 3rd International Conference
on Advanced Algorithms and Control Engineering (ICAACE), volume 1570, pages
1–7. IOP Publishing, 2020.
[65] M. Tolba, H. Rezk, A. A. Z. Diab, and M. Al-Dhaifallah. A novel robust method-
ology based salp swarm algorithm for allocation and capacity of renewable dis-
tributed generators on distribution grids. Energies, 11(10):2556, 2018.

[66] A. Viktorin, R. Senkerik, M. Pluhacek, T. Kadavy, and A. Zamuda. DISH-XX

Solving CEC2020 Single Objective Bound Constrained Numerical optimization
Benchmark. In Proc. Congress on Evolutionary Computation (CEC), pages 1–8.
IEEE, 2020.

21
This is a post-print version of an article published in Expert Systems with Applications. The final publication is available at
Elsevier via https://doi.org/10.1016/j.eswa.2021.116029

[67] C. C. Villalón, T. Stützle, and M. Dorigo. Cuckoo search≡(µ+ λ)–evolution

strategy. IRIDIA – Technical Report Series, 6:1–15, 2021.
[68] C. L. Villalón, M. Dorigo, and T. Stützle. The intelligent water drops algorithm:
why it cannot be considered a novel algorithm. Swarm Intelligence, 13(3):173–192,
2019.
[69] D. Weyland. A critical analysis of the harmony search algorithm—how not to
solve sudoku. Operations Research Perspectives, 2:97–105, 2015.

[70] F. Wilcoxon. Individual comparisons by ranking methods. In Breakthroughs in

statistics, pages 196–202. Springer, 1992.
[71] D. H. Wolpert and W. G. Macready. No free lunch theorems for optimization.
IEEE Trans. Evol. Comput., 1(1):67–82, 1997.
[72] B. Yang, L. Zhong, X. Zhang, H. Shu, T. Yu, H. Li, L. Jiang, and L. Sun. Novel
bio-inspired memetic salp swarm algorithm and application to MPPT for PV
systems considering partial shading condition. J. Clean. Prod., 215:1203–1222,
2019.
[73] W. Yang, J. Wang, H. Lu, T. Niu, and P. Du. Hybrid wind energy forecasting and
analysis system based on divide and conquer scheme: A case study in China. J.
Clean. Prod., 222:942–959, 2019.
[74] X.-S. Yang. A new metaheuristic bat-inspired algorithm. In Proc. Nature in-
spired cooperative strategies for optimization (NICSO), Studies in Computational
Intelligence, pages 65–74. Springer, 2010.
[75] X.-S. Yang and S. Deb. Engineering optimisation by cuckoo search. International
Journal of Mathematical Modelling and Numerical Optimisation, 1(4):330–343,
2010.
[76] X.-S. Yang and X. He. Firefly algorithm: recent advances and applications. Int. J.
Swarm Intell., 1(1):36–50, 2013.

22