www.nature.

com/scientificreports

OPEN Dynamic Bayesian network

structure learning based
on an improved bacterial foraging
optimization algorithm
Guanglei Meng 1,2, Zelin Cong 1,2*, Tingting Li 2, Chenguang Wang 2, Mingzhe Zhou 1,2 &
Biao Wang 1,2
With the rapid development of artificial intelligence and data science, Dynamic Bayesian Network
(DBN), as an effective probabilistic graphical model, has been widely used in many engineering
fields. And swarm intelligence algorithm is an optimization algorithm based on natural selection
with the characteristics of distributed, self-organization and robustness. By applying the high-
performance swarm intelligence algorithm to DBN structure learning, we can fully utilize the
algorithm’s global search capability to effectively process time-based data, improve the efficiency
of network generation and the accuracy of network structure. This study proposes an improved
bacterial foraging optimization algorithm (IBFO-A) to solve the problems of random step size, limited
group communication, and the inability to maintain a balance between global and local searching.
The IBFO-A algorithm framework comprises four layers. First, population initialization is achieved
using a logistics-sine chaotic mapping strategy as the basis for global optimization. Second, the
activity strategy of a colony foraging trend is constructed by combining the exploration phase of
the Osprey optimization algorithm. Subsequently, the strategy of bacterial colony propagation is
improved using a "genetic" approach and the Multi-point crossover operator. Finally, the elimination-
dispersal activity strategy is employed to escape the local optimal solution. To solve the problem
of complex DBN learning structures due to the introduction of time information, a DBN structure
learning method called IBFO-D, which is based on the IBFO-A algorithm framework, is proposed.
IBFO-D determines the edge direction of the structure by combining the dynamic K2 scoring function,
the designed V-structure orientation rule, and the trend activity strategy. Then, according to the
improved reproductive activity strategy, the concept of "survival of the fittest" is applied to the
network candidate solution while maintaining species diversity. Finally, the global optimal network
structure with the highest score is obtained based on the elimination-dispersal activity strategy.
Multiple tests and comparison experiments were conducted on 10 sets of benchmark test functions,
two non-temporal and temporal data types, and six data samples of two benchmark 2T-BN networks
to evaluate and analyze the optimization performance and structure learning ability of the proposed
algorithm under various data types. The experimental results demonstrated that IBFO-A exhibits
good convergence, stability, and accuracy, whereas IBFO-D is an effective approach for learning DBN
structures from data and has practical value for engineering applications.

Keywords Dynamic Bayesian networks, Structural learning, Swarm intelligence optimization algorithm,
Bacterial foraging optimization algorithm

Bayesian networks (BNs) combine probability theory with graph theory and exhibit strong interpretability and
high learning efficiency. The logical relationships inherent in data can be effectively investigated, offering a
promising technical approach for establishing causal models for complex p ­ roblems1. Dynamic Bayesian networks
(DBNs) are powerful algorithmic tools that integrate the structure of static BNs with time-related information and
are employed for dynamic uncertainty inference and temporal data analysis. DBNs have applications in various

1
School of Automation, Shenyang Aerospace University, Shenyang 110136, China. 2Aviation Science and
Technology Key Laboratory of Air Combat System Technology, Shenyang 110136, China. *email: congzelin@
stu.sau.edu.cn

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

fields, including artificial intelligence, machine learning, and automatic ­control2. Furthermore, DBNs have a
broad range of engineering applications, such as managing transcriptional regulatory relationships between
cancer ­genes3, identifying connectivity issues between human brain regions through high-order DBNs using
functional magnetic resonance imaging time series ­data4, and analyzing the vascularization in the formation
process of engineered tissues, aiming to enhance the accuracy of predicting future time steps and ensuring an
acceptable uncertainty in forecasting the future progress of the o ­ rganization5. Integrating structural prediction
methods, such as mutual information and maximum information coefficient into the DBN model enhances the
efficiency and scale of gene regulatory network r­ econstruction6.
However, the introduction of time information into the DBN network increases search space complexity,
reduces structure learning accuracy, and impedes the direct application of the static BN learning method. Early
learning approaches involved constructing DBN network structures by experts based on their prior e­ xperience7,8.
Conversely, for large datasets with numerous parameters and high complexity, relying only on expert experience
poses challenges in establishing causal relationships based on temporal information and ensuring structural
accuracy. To solve these problems, Leray et al.9 proposed a classical DBN structure learning algorithm inspired
by the dynamic max–min hill climb (DMMHC) local search algorithm. This algorithm allows for the quick
identification of constraints based on temporal information in the 2T-BN model. Trabelsi et al.10 introduced the
heuristic greedy search (GS) a­ lgorithm11, which was extended to 2T-BN networks to perform local optimal selec-
tion through a generative neighborhood algorithm as a way to expect to find the globally optimal DBN structure.
However, these classical DBN structure learning methods still have problems such as low efficiency of gen-
erating networks and lack of escaping local optimum mechanism. Therefore, how to further optimise the DBN
structure learning methods to improve their search efficiency and search accuracy is the current research hot-
spot for scholars. In recent years, researchers have shown significant interest in nature-inspired metaheuristic
algorithms based on swarm intelligence (SI) for optimization. Notable optimization algorithms include genetic
algorithm (GA)12, ant colony ­optimization13, particle swarm optimization (PSO)14,15, grey wolf optimization
(GWO)16, artificial bee ­colony17, bat ­optimization18, whale ­optimization19, and firefly ­optimization20. Scholars
have also improved the aspects of population initialization and individual optimization iterative updating strate-
gies of these optimization algorithms on the original basis by introducing a series of effective methods such as
Lévy ­flights21, opposition-based l­ earining22, adaptive strategies, and hybrid algorithms, aiming to improve their
optimization performance, as well as the accuracy and stability of the optimization results. Wu et al.23 proposed
an improved ant colony optimization algorithm (ICMPACO) by introducing multiple swarm strategies, co-
evolutionary mechanism, pheromone updating strategy and pheromone diffusion mechanism, which improved
the optimization ability and stability of the algorithm. And it was applied to Traveling Salesmen Problem to obtain
better allocation results. Gao et al.24 proposed an improved variable weight gray wolf optimization algorithm
VW-GWO, which improves the probability of the algorithm escaping from the local optimum by introducing
control parameters and decreasing control equations. Liu et al.25 proposed an adaptive weighted particle swarm
optimization algorithm (AWPSO) based on sigmoid function, which updated the acceleration coefficients using
the sigmoid activation function of the neural network and considered both the particles to the global optimal
position through an adaptive weighting strategy, thus improving the optimization efficiency and convergence
characteristics of the algorithm. Zhang et al.26 proposed a Chaotic Bacterial Foraging Optimization (ChaoticBFO)
algorithm to achieve a reasonable balance between exploration and exploitation by introducing a chaotic ini-
tialization strategy and a chaotic local search with a "contraction" strategy in the convergence step. Mou et al.27
proposed an adaptive non-dominated sorting genetic algorithm III (ANSGAIII), the algorithm enhances the
objective function by considering non-linear relationships, equality constraints, actuator rate and position con-
straints. The algorithm solves the problem of autonomous berthing and dynamic positioning of over-actuated
ships. Giri et al.28 proposed an adaptive neighbourhood for locally and globally tuned biogeography based opti-
mization algorithm (ANLGBBO) which inherits features of the nearest neighbour of the local best individual
to be migrated along with a global best individual of the pool. And explore large Spaces by identifying areas
with high-quality solutions. In addition, literature research shows that advanced optimization algorithms and
their variants have played an important role in practical engineering applications in many intelligent fields, such
as: Data mining e­ ngineering29,30, automatic d ­ riving31,32, intelligent r­ obots33, network topology a­ rchitecture34,35,
36,37
military intelligent e­ quipment and other fields.
The use of metaheuristic search mechanisms to investigate the solution space is a common characteristic
among SI-based algorithms. During each computation iteration, these algorithms update the state of the local
models and generate new populations to search for local optima. The quality of the solutions continuously
improves as the number of iterations increases, ultimately approaching the optimal model. Based on this, opti-
mizing the DBN network structure using a high-performance swarm intelligence optimization algorithm is a
novel and effective research method. Li et al.38 introduced a binary PSO algorithm based on mutual information,
which effectively prunes the search space. The algorithm accelerates the convergence speed of deep belief net-
work structure learning by updating the particles using a probability threshold. Heng et al.39 proposed a fitness
function and redefined the encoding format of the particle swarm algorithm for DBN structure learning from
incomplete datasets. Santos et al.40 redefined the positions and velocities in the particle swarm algorithm to learn
the high-order dynamic Bayesian network structure from large-scale and multivariate time series data. Jiang
et al.41 proposed an adaptive learning algorithm for DBN structure learning using a two-step strategy and adap-
tive crossover and mutation rates within a GA framework. Quesada et al.42 proposed an order-invariant, Markov
order-independent high-order particle encoding approach that exhibits scalability in high-order networks. Deng
et al.43 proposed an improved binary bat algorithm for learning the transitional network structure of a DBN and
constructed a fitness function to quantitatively assess the node order in the network.
Although the aforementioned SI-based DBN structure learning approaches have achieved a certain level of
effectiveness, there is still substantial potential for improvement in terms of optimization efficiency and accuracy.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 2

Vol:.(1234567890)
www.nature.com/scientificreports/

Passino et al.44 proposed a bacterial foraging optimization (BFO) algorithm that stimulates the foraging behavior
of Escherichia coli bacteria in the human body. BFO is a global stochastic search algorithm, The simulation of
the bacterial population comprises four steps: chemotaxis, grouping, reproduction, and elimination-dispersal.
It has the characteristics of not requiring the gradient information of the optimization object during the opti-
mization process, low complexity and fast convergence, which can be applied to reduce the number of iterative
convergence times for finding the candidate network, jumping out of the local optimum, and searching for the
highest scoring globally optimal DBN network structure. However, the original BFO also has certain defects,
such as random steps of chemotactic activity, poor information exchangeability of clustering mechanism, and
inability to maintain a balance between global and local search.
In summary, to enhance the BFO optimization performance, we propose a new hybrid algorithm, called the
improved bacterial foraging optimization algorithm (IBFO-A), which aims to improve optimization iteration
speed and accuracy while maintaining the low time complexity and fast convergence performance of BFO and
balancing the global and local exploration and development capabilities. Then, within the framework of the
IBFO-A algorithm, combined with a dynamic K2 scoring function and customized learning strategy, an IBFO-D
method for DBN structure learning is designed to improve its ability to optimize the learning network structure
from the data.
The main contributions of this study can be summarized as follows:

(1) To improve the population quality, the population is initialized using a logistics-sine chaotic mapping
strategy. During the development and exploration phase of hybrid osprey optimization algorithm (OOA),
the chemotactic activity of bacteria was reconstructed, improving the ability of individual bacteria to rec-
ognize and move toward the optimal target fitness value.
(2) Based on the replication idea of GA and the Multi-point crossover operator, the reproduction steps were
reconstructed. This involves crossing the poor individual Xworst and fusing the better individual Xbest ,
thereby improving performance, increasing the species diversity of the flora, and escaping the local optimal
solution based on the elimination-dispersal operator.
(3) A dynamic K2 scoring function and V structure orientation rule are established. Combined with the IBFOA
framework, the cumulative health score is saved during the breeding stage to reduce the number of iterative
convergence in searching for candidate networks, and it is used in the elimination-dispersal stage to find
the globally optimal DBN network structure with the highest score.
(4) The effectiveness of the proposed algorithms (IBFO-A and IBFO-D) is evaluated and comparisons are made
with other mature algorithms through experimental tests on the benchmark data set. Statistical analyses
of the experimental results are conducted as follows:
• Firstly, the IBFO-A algorithm is compared with seven other optimization algorithms using 10 sets of dif-
ferent types of CEC2005 benchmark functions (unimodal, multimodal, and hybrid). These include three
original algorithms, two classical algorithms, and two recent advanced algorithms. Additionally, sensitivity
analysis experiments were conducted for the three parameters of IBFO-A. The experimental results indicate
that IBFO-A algorithm runs stably, ASR ranks first and has a certain competitiveness. Subsequently, we
conducted comparative experiments on 12 optimization algorithms, including IBFO-A, using the CEC2019
benchmarking functions and two real-world engineering optimization problems. Some novel as well as
improved optimization algorithms are included. The experimental results show that the IBFO-A algorithm
exhibits good optimization performance, indicating its potential in real engineering applications.
• Secondly, the B0 and B→ network structure learning capabilities of IBFO-D in non-temporal and temporal
data samples are investigated, revealing that the generated network structure can converge stably within a
high fitness value.
• Finally, IBFO-D is compared with two other structure learning algorithms using six 2T-BN temporal net-
work data samples. The experimental results show that IBFO-D is an effective method for optimizing DBN
structure learning from the data.

The remaining sections of this study are structured as follows. Preliminaries offers a review of the relevant
concepts and scoring metrics of BN, along with an introduction to the basics of DBN. In Methodology, the
principles of IBFO-A and IBFO-D algorithms and the design of dynamic scoring function are described in
detail. Experimental section presents the results and analysis of the simulation experiments. Finally, Conclusion
provides the conclusion and outlines plans for future research.

Preliminaries
Static Bayesian network
The Bayesian network N is represented as a binary tuple N = (G, �) comprising structure G and network param-
eters . In graph theory, the independent relationships among a set of variables can be represented using a
directed acyclic graph (DAG). Here, G = (X, E) represents a specific instance or representation of such a graph;
where X is a nonempty set of all nodes in the graph. X = {X1 , X2 , . . . , Xi , . . . , Xn }, Xi can be either an observed
variable or a latent variable; E is the set of directed line segments between different variables in the DAG, and
Xj → Xi d represents the direct dependencies between ­nodes45.

(1)
 
E = Xj → Xi |Xj ∈ pa(Xi ), i = 1, . . . , n

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

where pa(Xi ) is the "causes" of the node Xi , also called the set of parent nodes. Given the parent node set pa(Xi ),
Xi is independent of its non-descendant node set nd(Xi ) based on Markov independence. Thus, the joint prob-
ability of several nodes Xi that follow the Markov rule can be expressed as follows:
n

P(X1 , · · · , Xn ) = P(Xi |pa(Xi )) (2)
i=1

The conditional probability table of each node Xi given its known parent node set pa(Xi ) is represented by
the network parameter � = {�1 , �2 , . . . , �n }. It is possible to calculate the joint probability distribution of the
node Xi when the network structure G and the network parameters  of a Bayesian network are known. Com-
pared with other approaches for calculating joint probabilities, the efficiency of Bayesian network algorithms is
significantly higher because of the conditional independence among nodes.

Scoring function
The search and score-based BN learning approach mainly comprises two parts: model selection and model
optimization. Its core idea involves considering all possible structures as the domain, selecting a scoring func-
tion that assesses the quality of specific structures, and treating the process of identifying the best structure as
an optimization problem of searching for the optimal value of the scoring function within the domain.
Prior knowledge about structure G is summarized as a probability distribution P(G), referred to as the
structure prior distribution for a Bayesian network N = (G, �). Similarly, prior knowledge about parameters 
is summarized as another probability distribution P(�|G) referred to as the parameter prior distribution for a
given structure G . In this manner, the prior distribution of N can be expressed as follows:
P(G, �) = P(G)P(�|G) (3)
The posterior probability distribution P(G|D) is calculated when given an observed dataset
D = {D1 , D2 , . . . , DN }. Only the structural models G∗ corresponding to the maximum posterior probability
distribution in the search space are considered.
G∗ = arg max P(G|D) (4)
G

And
P(D|G)P(G)
P(G|D) = (5)
P(D)
Selecting the structure with the maximum posterior probability is equivalent to selecting the structure that
maximizes the following function since P(D) does not depend on G:
arg min log P(G|D) = arg min log P(D|G) + log P(G) + C (6)
G G

Based on penalized maximum likelihood or marginal likelihood, various scoring metrics, including Bayes-
ian Dirichlet, Bayesian Dirichlet equivalent, K2, minimum description length, Bayesian information criterion,
and mutual information test, have been proposed to assess the fitness of networks during the search process.
The most classic K2 scoring function formula is expressed as follows:
qi
n  ri
 (ri − 1)! 
P(G, D) = P(G)
(Nij + ri − 1)!
Nijk ! (7)
i=1 j=1 k=1

where n denotes the number of variables in the sample, qi denotes the number of parent nodes for Xi , ri denotes
the number of possible values for Xi , Nijk denotes the number of samples, and Nij denotes the total number of
samples.

Dynamic Bayesian networks

DBN is a graphical model structure that illustrates the conditional independence relationships between random
variables and their temporal evolution p ­ atterns46. Its unique transition network can reflect the state changes of
the system under different environmental factors in various time slices, showing the complex interactions and
dependencies among variables in the system and offering a closer approximation to the real situations of dynamic
multidimensional data.
However, representing X1 , X2 , . . . , Xn stochastic processes using DBN requires deriving a probability distribu-
tion over the random variable a, which can be highly complex. Thus, it is crucial to make appropriate assumptions
about DBN and design a reasonable and efficient optimization algorithm for structure learning to study and
model complex systems (see Methodology). These assumptions can be summarized as follows:

(1) The marginal directionality rule describes the dependency relationships between nodes in a finite time slice
t, and the changes in conditional probabilities tend to converge to consistent stability across all processes.
(2) Given the random variables at time step t, the random variables at time step t + 1 are conditionally inde-
pendent of the remaining random variables;Xt+1 ||(Xt−1 , X0 )|Xt . In other words, the Markov chain prop-

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 4

Vol:.(1234567890)
www.nature.com/scientificreports/

erty is satisfied to P(Xt+1 |X0 , X1 , . . . , Xt ) = P(Xt+1 |Xt ) by the entire dynamic discrete-time probabilistic
­process47.
(3) Across all adjacent time steps, the network topology remains invariant and the transitional network, along
with its corresponding conditional probability dependencies, remains the same. In other words, P(Xt+1 |Xt )
is independent of time t.

The DBN constructed on the time trajectory of the random process comprises two components based on the
aforementioned conditions: (B0 , B→ ).

(1) The initial network B0, defined on the initial state X0, and the joint probability distribution P(X0 ) obtained
from it form the most initial graphical structure of the Bayesian network (BN) from which the prior prob-
abilities of any node can be derived.
(2) The graphical structure of the BN composed of more than two time steps is represented by the transitional
network B→, defined by variables X0 and X1, with transitional probabilities P(Xt+1 |Xt ).

In other words, the entire DBN corresponds to {0, 1, 2, . . . , T} finite period, a and unfolds the probabilistic
graphical model onto the topology of the random variable X0 , X1 , . . . , XT . The parent nodes of X0 are those in
the initial network B0 at time 0. At time t + 1, the parent nodes of Xt+1 are those in the transitional network B→
that are relevant in both time steps t and t + 1. A set of initial networks B0, a transitional network B→, and a
simple DBN model structure with two time slices are illustrated in Fig. 1.
To summarize, given a DBN model, the joint probability distribution on X0 , X1 , . . . , XT is defined as follows:
T

P(X0 , X1 , . . . , XT ) = PB0 (X0 ) PB→ (Xt+1 |Xt ) (8)
t=0

Figure 1.  Dynamic Bayesian network model diagram.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

To solve the actual optimization decision problem, the SI-based structure learning method of DBN extends
the static optimization model, starting with the static initial network B0. This process involves the construction
of a basic graph model for dynamic intelligent optimization using time slice information. Figure 2 shows the
specific algorithmic process. A round of the BN network node set can be generated through the transitional
network t + 1 when the environmental factors change in round B→. A new population is generated along with
the actual optimization problem based on the BN nodes in the round t + 1, which is then evolved and optimized
to produce a set of excellent solutions and an optimal BN structure graph that matches the current environment,
serving as the most suitable reasoning tool for the current problem. Subsequently, the node set to be optimized
in the round t + 2 is generated by the DBN, and this process continues. As environmental factors change, infer-
ence and optimization are conducted to effectively address various emergencies and enhance the mitigation of
the effect of uncertain factors on the findings.

Methodology
IBFO‑A optimization algorithm
Population initialization
For swarm intelligent optimization algorithms, the selection of individual initial positions often affects the
algorithm’s iterative convergence performance. The original BFO algorithm uses random initial locations, which
results in the dispersion of most bacteria generated at the initial moment being far away and even not meeting
the boundary constraints. Chaotic ­mapping48 is an effective method to improve the population initialization of
the optimization algorithm. In this study, a logistics-sine mixed method proposed by Demir et al.49 is used to
integrate the two most universal methods of chaotic mapping: Logistic mapping and Sine mapping evenly dis-
tribute the population in the mapping space, significantly improving the species diversity and search efficiency
of the population in the following ways:
The upper bound of the feasible domain of each dimension of the objective function ub = [ub1 , ub2 , . . . , ubd ]
and the lower bound lb = [lb1 , lb2 , . . . , lbd ]. The location matrix modeling of bacterial individuals in the search
space is as follows:
x1,1 · · · x1,j · · · x1,m
 
 .. . . . . .. 
 .
 . .. . . . 
X=  xi,1 · · · xi,j · · · xi,m 
 (9)
 . . . . .
. . . . .

 . . . . . 
xn,1 · · · xn,j · · · xn,m n×m

Use the logistics-sine method to initialize the bacterial individual location:

ai+1 = rai (1 − ai ) (10)

Figure 2.  DBN intelligent optimization model diagram.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 6

Vol:.(1234567890)
www.nature.com/scientificreports/

(4 − r) sin (πbi )
bi+1 = (11)
4

xi+1 = (ai+1 + bi+1 )(mod1) (12)

where ai ∈ (0, 1) and bi ∈ (0, 1) are randomly generated series, ai+1 represents the logistic chaotic mapping, bi+1
represents the sine chaotic mapping, r represents the chaos coefficient, and xi+1 is the bacterial chaotic mapping
value determined by logistics-sine.
Finally, the chaotic sequence is mapped to the solution space:
Xi = lb + xi+1 (ub − lb) (13)

Chemotactic activity
Chemotactic activity plays a crucial role in the IBFO-A’s algorithm, in which bacteria first tentatively choose
the direction for a "flip" motion, and then swim to a nutrient-rich area through a "swim" motion. In the origi-
nal BFO algorithm, the trend activity is randomly given the i th bacterial movement step C(i) and receives the
attraction signal from other individuals in the population to swim to the center of the population, and the attrac-
tion between bacteria is represented by Jicc θ , θ i (j, k, 1) , i = 1, 2, . . . , S. At the same time, there will be repul-
sion between bacteria, which prevents the consumption of nearby nutrients by maintaining a certain distance.
Jcc (θ, P(j, k, l)) said that the combined influence of attraction and repulsion between bacteria is considered at
the same time, and its computation formula is as follows:
s

i
Jcc (θ, P(j, k, l)) = JCC (θ, θ i (j, k, l))
i=1
s
  p 
i 2
   
= −datact exp −wattact θm − θm (14)
i=1 m=1
s
  p 
i 2
   
+ hrepenatint exp −wrepelinat θm − θm
i=1 m=1

where P(j, k, l) represents the position of each bacterium in the population S after the j trend operation, the k
replication operation, and the l elimination-dispersal operation, θ = [θ1 , . . . , θD ]T is a point on the optimization
domain, θmi is the m element of the i bacteria, dattract represents the amount of attraction released by the bacte-
ria, Wattract is used to measure the width of the attraction signal. hrepenatint indicates the amount of rejection
released by the bacteria, and Wrepelinat measures the width of the rejection signal.
However, there are some problems in the original BFO trend activity. First, the bacterial movement step C(i)
is given randomly, resulting in a low convergence accuracy of the algorithm. To solve this problem, most scholars
choose to design a new step size. Supriyono et al.50 developed three types of step size strategies: linear step size,
quadratic step size, and exponential step size. Niu et al.51 proposed a linear chemotactic decline step and a non-
linear chemotactic decline step as well as other types of non-adaptive ­steps52–54.
In addition, the effect of communication between bacterial groups is limited, but the clustering mechanism
with complex objective function cannot effectively guide bacterial individuals to the high-nutrient (fitness value)
region, resulting in the algorithm often falling into the local optimal value prematurely. To solve this problem,
scholars often choose to ignore the original clustering mechanism and combine better communication mecha-
nisms to improve the algorithm. Chen et al.55 combined the PSO algorithm to enhance intercellular communica-
tion and proposed an adaptive foraging strategy using area-focused search. Wang et al.56 also chose to combine
the PSO algorithm and Gaussian distribution to adjust the chemotactic activity of the flora and strengthen the
ability of information exchange among the populations. Zhao et al.57 employed the gravitational mechanism in
GSA to improve the ability of information exchange between individuals in the chemotactic step of the BFO
algorithm.
The Osprey optimization algorithm was proposed by Mohammad Dehghani and Pavel Trojovsky in 2023 to
simulate the predation behavior of ­Osprey58. In the first stage of OOA, the Osprey identifies the position of the
fish (fitness value) and performs the arrest (moving in the direction of high fitness and updating the individual
position). For each Osprey, the position of the other Osprey with a better target fitness value in the search space
is also regarded as the fish school. OOA Phase 2 brings the fish to the appropriate position to feed (moving in a
random direction and updating the individual position). Among them, FPi operator with certain clustering and
optimal value searching ability in the OOA algorithm, and xi,j P1 operator with more optimal positions to update
individual positions, can be used to solve the problems in the BFO algorithm. In summary, this study chose to
combine the first phase of OOA with BFO chemotaxis to fully improve the performance of the IBFO-A algorithm.
The second stage and subsequent elimination-dispersal activities are not selected to be combined with this stage.
The mathematical formula of "flip" movement:
FPi = {Xk |k ∈ {1, 2, . . . , N} ∧ Fk < Fi } ∪ {Xbest } (15)
Formula 15 is used to investigate the search space with a good target value, where FP is a set of i target loca-
tions and XB is the best candidate solution.
The mathematical formula of "swimming":

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

P1
 
xi,j = xi,j + ri,j · SFi,j − Ii,j · xi,j (16)


XiP1 , FiP1 < Fi
Xi =
Xi , else (17)

Formula 16 calculates the new position of the individual, and if this new position improves the target fit-
ness value, the previous position is replaced according to Formula 17, where Xi is the original position of the
P1 is the new position of the individual i, F P1 is its objective function value, SF is the candidate
individual, xi,j i i,j
solution chosen by the individual i, ri,j is the random number in the interval [0,1], and Ii,j is the random number
in the set {1, 2}.

Reproductive activity
Reproduction is a fundamental biological behavior observed in various species and is a crucial aspect of life
preservation. With each reproductive generation, the search efficiency of the colony improves. It has been men-
tioned in the conventional BFO algorithm that inferior bacteria should be eliminated and superior bacteria
retained for reproduction. The cumulative state of the health fitness value of the i bacteria is represented by
Nc+1
Ji = J i, j, k, l .
  
j=1
The population of bacteria is divided into two sets based on the accumulation of their health fitness scores:
Xbest , comprising the top-ranked elite individuals with higher cumulative scores and Xworst , comprising the
lower-ranked inferior individuals with lower cumulative scores. Then, a “genetic” approach was employed to
perform reproduction activities following the specific formula below:
′
Xbest = Xbest (18)

′
Xco = Xbest ⊗ Xworst (19)

X ′ = Xco ⊕ Xbest (20)

59
where ⊗ represents a crossover operator employed to perform the "Multi-point crossover" of the encoding.⊕
denotes the fusion operator. Xbest
′ represents the reproduction elite individual obtained by X
best after replication
and coding. Xbest
′ and Xworst perform crossover operations on their encodings, leading to a single reproductive
crossover individual Xco. X ′ denotes the new bacterial population obtained by fusing with Xco and Xbest . This
method enables an increase in the chemotactic ability of bacteria with lower cumulative scores, i.e., it enhances
the average quality of the entire population while maintaining the original total number of bacterial individuals
S. Furthermore, it improves the species diversity of the population and prevents the algorithm from becoming
trapped in local optima.

Elimination‑dispersal activity
In this study, the elimination-dispersal activity mechanism was enhanced based on adaptation theory. The bacte-
rial population randomly selects and performs elimination-dispersal activity after each Nr round of reproduction,
prompting the bacterial individuals to produce new solutions and conduct a new search for positions, thereby
escaping local optima.
The specific definition of the elimination-dispersal function is expressed as follows:
f  + (f − f ) ∗ P i−1 
   
i min i
(21)
max ed
Ped =    
fmin  + (fmax − fmin ) ∗ Pedi−1 

where Pedi is the elimination-dispersal probability at the current moment Pedi−1 represents the probability of
elimination-dispersal at the previous time. fmin represents the worst goal score in history, fmax represents the best
goal score in history, and fi represents the current goal score of specific bacteria i.
With an increase in the number of iterations, the adaptive elimination-dispersal probability shows a non-
linear decreasing trend. At the early stage of iteration, to explore the solution space more widely, a larger elim-
ination-dispersal probability is needed to find other foraging paths. In the later iteration, due to the guidance
of the global optimal solution, the algorithm conducts a fine search near the global optimal solution, and the
elimination-dispersal probability is reduced. Thus, the local development ability is enhanced, and the bacteria
can find the target solution more quickly and accurately. In addition, for formula (21), to determine a better
objective function, the population elimination-dispersal probability increases when the current score is close
to the lowest score.

K2 dynamic scoring function

K2 scoring function differs from that of static BN because of the introduction of time information in DBN. A
dynamic scoring function is necessary to measure the validity of the network structure. Thus, the K2 dynamic
scoring function in IBFO-D is discussed in this section.
First, the initial network B0 can be learned from the dataset assuming that the training set consists of N
complete sequence samples, where the length of the l -th sample is Nl , and a specific value is assigned to the

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 8

Vol:.(1234567890)
www.nature.com/scientificreports/

variable x l [0], . . . , x l [Nl ]. Then, the transitional network B→ can be learned from the transformed data N = Nl .

Considering the definitions of network parameters t = 1, · · · , T and sequence samples θ within the time slice N :
0 = P X [0] = k |Pa X [0] = j
  
θi,j,k i i i i
→ = P X [t] = k |Pa(X [t]) = j
θi,j,k
  (22)
k i k i

 0
I Xi [0] = ki , Pa Xi [0] = ji ; x l
� � � �
 Ni,j,k =
�l � � (23)
→ = I Xi [t] = ki , Pa(Xi [t]) = ji ; x l
�
 Ni,j,k
l 0

where I represents an indicator function, specifically defined as:

  
1 if Xi [t] = ki , Pa(Xi [t]) = ji
I=
0 else (24)

The joint probability density of DBN is expressed as follows:

T−1

PDBN (x[0], x[1], . . . , x[T]) = PB0 (X[0]) PB→ (x[t + 1]|x[t]) (25)
t=0

The structure of DBN decomposes the likelihood function distribution into:


P(D|G) = P(D|G, θ )P(θ|G)dθ (26)

The first term of the integral is decomposed into the following formula:
 N 0     N →
0 i,j,k
P(D|G, θ ) = θi,j,k · θi,j,k i,j,k (27)
i j k i j k

The second term of the integral is decomposed into the following formula, assuming that the prior distribu-
tion on conditional probabilities is conditionally independent:
     
0 →
P(θ|G) = P θi,j,k · P θi,j,k (28)
i j i j

The likelihood function can be rewritten as the product of two integrals by substituting the aforementioned
formula:
  N 0  
0 i,j,k 0 0
P(D|G) = θi,j,k × P θi,j,k × dθi,j,k (29)
i j k

The likelihood function P(D|G) can be further expressed as a product of K20 and K2→ given the hyperpa-
rameter Ni,j,k
0 N → and complete data, and with the parameter prior following a Dirichlet distribution:
i,j,k
qi
n  ri
 (ri − 1)! 
→
P(D|G) =
(Ni,j0 + ri − 1)!
· Ni,j,k ! (30)
i=1 j=1 k=1

where n denotes the number of variable samples, qi denotes the number of parent nodes for Xi , ri denotes the
number of possible values for Xi , Ni,j0 denotes the number of samples, and Ni,j,k
→ denotes the total number of

samples.
This study selects the logarithm of the likelihood function to minimize in practical applications. The final
expression of the dynamic K2 scoring formula is as follows:
Score(G|D) = K20 (G|D) + K2→ (G|D)
n qi ri
(31)
      
= qi · log (ri − 1)! − log Ni,j0 + ri − 1 ! + →
log Ni,j,k !
i=1 j=1 k=1

IBFO‑D optimization algorithm

Initialization
In Swarm intelligent optimization algorithms, encoding approaches to generate abstract structures and the
concretization of the optimization process are crucial elements. In this study, the network structure is repre-
sented using an adjacency matrix A = (aij ) with n × n dimensions. The directed edge from node i to node j is
represented by aij = 1, whereas the absence of a connection between node i and node j is denoted by aij = 0.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 9

Vol.:(0123456789)
 www.nature.com/scientificreports/


1 i is a parent of j
aij =
0 no edges or deleted edges (32)

The specific process first initializes an empty adjacency matrix, develops and explores the search space through
bacteria in the direction of high fitness, and constantly updates the location if and only if the new location has a
higher score. Then, the network graph structure is updated by directional rules, and the process is repeated until
the network structure with the highest K2 score in the time slice t is found, and the optimization iteration of the
t + 1 time slice is started. The DBN structure is represented as a DAG. Thus, it is crucial to consider their validity
when constructing the initial network. In other words, in the searched DBN, the network structure should not
contain any cycles or bidirectional edges. Reflected in the adjacency matrix, this indicates that the nodes in the
matrix should not form cycles and that elements symmetrically located about the diagonal should not be to 1.
Figures 3 and 4 illustrate examples of generating initial network B0 and transfer network B→ structural adjacency
matrices, respectively. Node labels can be simplified to further reduce the search space, leading to B0 = 11|01|00
and B→ = 00110|00011|00001|00010|00001|00000.

Network structure learning

IBFO-D In the framework of the IBFO-A algorithm, the state relationship between nodes in the DBN was con-
sidered. Restrictions were imposed on the tendency directions of the bacteria, such as NS and NC , considering
the state relationships between nodes in the DBN. Consequently, three edge orientation rules, namely “add edge,”
“remove edge,” and “reverse edge,” were designed:

(1) Add Edge: Given a collection of nonempty nodes X = x1 , x2 , . . . , xi , xi+1 , . . . , xj |xi ∈ G, xi ∈ / (xj ) , if
  

eij = xi → xj is added and eij ∈ G holds, then G′ ∈ G ∪ (eij );

(2) Remove Edge: Given a directed edge set E = eij = xi → xj |xi ∈ (xj ), i = 1, · · · , n , if eij is removed,
  

then G′ ∈ G\(eij ) X = x1 , x2 , . . . , xi , xi+1 , . . . , xj |xi ∈ G, xi ∈

/ (xj ) “ and eij = xi → xj are removed;
  

(3) Reverse Edge: Given a directed edge set E = eij = xi → xj |xi ∈ (xj ), i = 1, . . . , n , if eij is removed,
  

eji = xj → xi is added, and eji ∈ G holds, then G′ ∈ G\(eij ) ∪ (eji ).

Figure 3.  Initial network B0.

Figure 4.  Transfer network B→.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 10

Vol:.(1234567890)
www.nature.com/scientificreports/

The chemotactic activity process continued until the bacteria reached a fixed position and no longer moved
or had moved the maximum number of chemotaxis, which corresponded to finding the network structure with
the highest K2 score in the DBN network or reaching the upper limit of search iterations.
To choose healthy bacteria, it is essential to assess the health level of each bacterium. This assessment involves
computing the sum of the fitness values of the chemotaxis steps. A higher cumulative value signifies that maxi-
mum nutrition has been obtained, making it more suitable for reproduction. In this study, the step fitness value
of bacteria is defined based on the dynamic scoring function in DBN structure learning. Specifically, the health
score for the i bacterium is expressed as follows:
Nc

HSti = K20(G|D) (i, j, k, l) + K2→(G|D) (it , jt , kt , lt ) (33)
jt =1

K20(G|D) (i, j, k, l) is defined as the prior fitness value function for the i bacterium during the j chemotaxis, k
reproduction, and l elimination-dispersal when generating the initial network. K2→(G|D) (it , jt , kt , lt ) is the fit-
ness value function for subsequent transition networks. Health function assesses the accumulated K2 score for
individual bacteria throughout the entire process of chemotaxis operations.
To determine the global optimal network structure, the specific definition of the elimination-dispersal func-
tion is expressed as follows:
HS  + (HS − HSi ) ∗ P t−1 
   
min t
Pedt =    max ed  (34)
HS  + (HS − HS ) ∗ P t−1 
min max min ed

where P ted
is the elimination-dispersal probability at the current moment. Pedt−1 is the elimination-dispersal prob-
ability of the previous moment. HSmin is the lowest historical health score, HSmax is the highest historical health
score, and HSti is the current specific health score of bacterium i.

Algorithm description
The IBFO-D Algorithm proposed in this study is shown in Algorithm 1. The whole DBN structure learning
process is summarized as follows: the algorithm starts from the initialization of network parameters, randomly
generates the initial DAG population, and finds the high-quality network structure through the chemotactic
activity formula (15–17). At the same time, the driving force of DBN local optimization is generated according
to three operators. The fitness value of each bacterium, namely the K2 score, was calculated, and the cumulative
value HSti was recorded as a health score. By selecting elite individuals with high HSti , the average optimization
ability of the bacterial population was updated according to the formula (18–20) to improve the convergence
speed, while preserving certain species diversity to prevent falling into local optimality. According to formula
(34), determine whether the bacterial individual generates a new solution and re-searches. According to the
above optimization steps, as well as the dynamic K2 scoring measures and constraints, until a high-score network
structure matching the data set is searched.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 11

Vol.:(0123456789)
 www.nature.com/scientificreports/

Algorithm 1.  IBFO-D

Experimental section
Experimental preparation
In the test experiment, we mainly focus on the optimization performance of the proposed method and its learning
effect on the DBN model structure, and do not deny the validity and novelty of other optimization algorithms
and their modeling in the corresponding domain. The parameter values of each algorithm used in the experi-
ment are shown in Table 1.
To test the optimization performance of the proposed algorithm, Firstly, we use IBFO-A and seven other
optimization algorithms to perform comparative test experiments on 10 different benchmark functions. The
reference function comes from ­CEC200560, including the multi-peak function, single-peak function, and fixed-
dimensional multi-peak function, which is used to test the convergence speed, accuracy, effectiveness, and global
search ability of the algorithm. The specific reference function is shown in Table 2. Then we select three kinds of
hyperparameters for parameter sensitivity analysis to test the stability of the algorithm.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 12

Vol:.(1234567890)
www.nature.com/scientificreports/

Secondly, We conducted several more sets of comparative experiments of optimization algorithms. These
include the eight optimization algorithms mentioned above and some novel optimization algorithms as well as
improved optimization algorithms. The benchmark function for the simulation experiments is from CEC2019,
which is a popular benchmark function that is quite effective in testing the performance of optimization algo-
rithms. Then we use 2 sets of real-world engineering optimization problems to test the optimization performance
of the algorithms from multiple perspectives. The CEC2019 specific reference function is shown in Table 3.
Finally, for the IBFO-D algorithm network learning performance test, various dynamic benchmark network
experiments were selected. These dynamic benchmark networks were derived from two well-known static BN
networks: the Asia and Alarm networks. These 2T-BN networks comprise two time slices and include an initial
network represented as B0 and a transition network represented as B→. Benchmark network data can be obtained
from the provided Supplementary URL online.

Method Population size Maximum generation Other parameters

IBFO-A 60 1000 r = 0.86; Ped = 0.8; pc = 0.2
BFO 60 1000 Ped = 0.8;  ∈ [−1, 1]
OOA 60 1000 r ∈ [0, 1]; I = {1, 2}
GA 60 1000 pc = 0.8; pm = 0.05
GWO 60 1000 a ∈ [2, 0]; |r1 | ∈ [0, 1]; |r2 | ∈ [0, 1]
PSO 60 1000 wmax = 0.9; wmin = 0.6; vmax = 6; c1 = 2; c2 = 2
BWO 60 1000 r1 ∈ (0, 1); r2 ∈ (0, 1); β = 1.5
DBO 60 1000 k ∈ (0, 0.2]; b ∈ (0, 1)
COA 60 1000 temp = rand ∗ 15 + 20
GO 60 1000 pro ∈ (0, 1); rnd ∈ (0, 1); coe ∈ (0, 1)
AWPSO 60 1000 wmax = 0.9; wmin = 0.4; vmax = 0.5; vmin = −0.5
ChaoticBFO 60 1000 µ = 0.8; Ped = 0.8;  ∈ [−1, 1]

Table 1.  Parameter settings for the algorithms used.

Test function Category D S fmin

n
xi2 30 0

f1 (x) = [−100, 100]n
i=1
n n
30 0
 
f2 (x) = |xi | + |xi |
Unimodal test functions [−10, 10]n
i=1 i=1
 2
n i
30 [−100, 100]n 0
 
f3 (x) = xj
i=1 j=1

n 
xi2 − 10 cos (2π xi ) + 10

30 0

f4 (x) = [−5.12, 5.12]n
i=1
n n  
1 xi
xi2 − 30 0
 
f5 (x) = 4000 cos √
i
+1 [−600, 600]n
i=1 i=1
 n−1
 Multimodal test functions
f6 (x) = 0.1 sin2 (3π x1 ) + (xi − 1)2 1 + sin2 (3π xi+1 )
 

i=1
n
30 [−50, 50]n 0
 
+(xn − 1) 1 + sin2 (2π xn ) +

u(xi , 5, 100, 4)
i=1
11  x1 bi2 +bi x2 2
 
4 [−5, 5]n 3.075e−4

f7 (x) = ai − bi2 +bi x3 +x4
i=1
 2 
f8 (x) = 1 + x1 + x2 + 1 19 − 14x1 + 3x12 − 14x2


+6x1 x2 + 3x22 × 30 + (2x1 − 3x2 )2 (18 − 32x1 2 [−2, 2]n 3

 

+12x12 + 48x2 − 36x1 x2 + 27x22 Multimodal test functions with fix dimension


 
4 6 2
6 − 3.32

[0, 1]n
 
f9 (x) = − ci exp − aij xj − pij
i=1 j=1

10  −1
(x − ai )(x − ai )T + ci 4 [0, 10]n − 10

f10 (x) = −
i=1

Table 2.  CEC2005 benchmark functions used in the experimental study. where D is the dimension of the
function, fmin is the minimum value of the function, and search range S ⊆ Rn.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 13

Vol.:(0123456789)
 www.nature.com/scientificreports/

The experimental setup comprised the following environment: Windows 11 operating system, MATLAB
and Python programming language, 32.0 GB RAM, an Intel Core i7-12700 K CPU running at 5.0 GHz, and an
NVIDIA GeForce RTX 3080Ti graphics card.

Experimental results and analysis of IBFO‑A compared with other SI algorithms

Comparison and analysis of eight optimization algorithms in CEC2005 benchmark functions
The experimental design of algorithm optimization performance comparison is as follows: The optimization
algorithm in IBFO-A is compared with three original optimization algorithms, BFO, OOA, and GA; two classical
optimization algorithms, GWO and PSO, two recent advanced optimization algorithms B ­ WO61 and D ­ BO62, and
a total of eight optimization algorithms are compared and tested on 10 groups of different types of benchmark
functions. We uniformly set experimental parameters for all optimization algorithms, in which the population
size is set to N = 60, the maximum number of iterations T = 1000, and the upper bound ub, lower bound lb,
dimension D , and optimal value fmin of different test functions are set as shown in Table 2. We present 10 sets of
optimization convergence curves, specific scores, and running timelines, and record the average score ranking
(ASR) and average run-time ranking (ATR) of the algorithm that runs 30 times independently (if both algorithms
converge to the optimal value, the ASR is determined by the number of iterations).
The simulation results in Table 4 show that IBFO-A can converge to the optimal value for 6 of the 10 bench-
mark functions. In unimodal and multimodal functions, IBFO-A converges directly to the optimal value 0 on the
F1, F2, F3, F4, and F5 functions. In addition, it can converge directly to the optimal value 3 on the F8 function,
and it is also very close to the theoretical optimal value in other fixed-dimensional multi-peak test functions,
which shows that it has good global optimization ability.
Here, we choose the original BFO, OOA, and GA algorithms as reference objects. According to the analysis
in Fig. 5, the GA and BFO algorithms perform poorly on F1, F2, and F3 unimodal functions, BFO improves
somewhat on multi-modal functions F4 and F5, and the OOA algorithm performs better. IBFO-A can stably
converge to the optimal with less than half of the OOA iterations. GA still performs poorly in F7 and F9 functions
of fixed dimension, and the convergence values of OOA and BFO are also different from the theoretical optimal
values. Compared with the IBFO-A algorithm, the performance of the IBFO-A algorithm is competitive. From
the experimental data in Table 4 and the convergence curve in Fig. 5, except for poor performance on the F6
generalized penalized function, compared with the other seven algorithms, IBFO-A exhibits the best performance
in the seven function scoring tests, and ASR ranks first. This shows that the IBFO-A algorithm has good conver-
gence speed and accuracy, and proves that the improved chemotactic step and the replication step using cross
strategy can avoid falling into the local optimal solution and enhance the local search ability of the algorithm.
In addition, the ATR of the IBFO-A algorithm ranks fourth. From the perspective of algorithm time complexity,
the time complexity of IBFO-A and BFO is O(n), whereas that of OOA is O(n2 ). Therefore, the computation time
of IBFO-A is better than that of OOA. However, due to the extra computing steps, it consumes more computing
time than the classical BFO, PSO, and GWO, which is also a limitation of the algorithm in this study.

Sensitivity analysis
This section discusses the hyperparameter sensitivity analysis of IBFO-A algorithm. We selected three hyper-
parameters that mainly affect the optimization performance of IBFO-A algorithm, including population size N,
elimination-dispersal probability Ped , and crossover probability pc . We set four different parameter values for
each hyperparameter to be discussed, and use the IBFO-A algorithm to optimize several typical test functions
of CEC2005 under these parameter settings. The specific parameter values and results are shown in Figs. 6, 7,
8 and Table 5.
From the sensitivity analysis of IBFO-A to hyperparameter N , it can be seen that with the increase of popu-
lation size, the probability of finding the global optimal solution will increase, thus improving the optimiza-
tion performance of the algorithm. However, large population sizes can also lead to increased time costs. Ped
determines the probability of initial elimination-dispersal occurrence of individual bacteria. According to the
sensitivity analysis of IBFOA to hyperparameter Ped , high Ped parameter value enables bacteria to explore the

Test function Description D S fmin

F1 Storn’s Chebyshev polynomial fitting problem 9 [−8192, 8192] 1
F2 Inverse Hilbert matrix problem 16 [−16384, 16384] 1
F3 Lennard–Jones minimum energy cluster 18 [−4, 4] 1
F4 Rastrigin’s function 10 [−100, 100] 1
F5 Grienwank’s function 10 [−100, 100] 1
F6 Weierstrass funetion 10 [−100, 100] 1
F7 Modified Schwefel’s function 10 [−100, 100] 1
F8 Expanded Schaffer’s F6 function 10 [−100, 100] 1
F9 Happy Cat function 10 [−100, 100] 1
F10 Ackley function 10 [−100, 100] 1

Table 3.  CEC2019 benchmark functions used in the experimental study.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 14

Vol:.(1234567890)
 Scientific Reports |
(2024) 14:8266 |
www.nature.com/scientificreports/

IBFO-A BFO OOA GA GWO PSO BWO DBO

https://doi.org/10.1038/s41598-024-58806-0
Score Time Score Time Score Time Score Time Score Time Score Time Score Time Score Time
F1 0 0.34518 3.8826e−4 0.304 0 0.37524 7.2368e−03 0.35948 1.1303e−152 0.27802 1.1153e−10 0.11689 0 0.49239 0 0.39853
F2 0 0.35709 1.3612 0.26638 0 0.37985 0.025851 0.38608 3.2631e−84 0.27342 3.9462e−06 0.13724 1.0557e−245 0.60831 5.7691e−179 0.4057
F3 0 0.50364 0.19572 0.26274 0 0.74867 3692.011 0.605 7.4765e−78 0.40169 7.6957e−05 0.2619 0 0.65241 0 0.80969
F4 0 0.306 0 0.24253 0 0.31499 3.0364 0.35543 0 0.29823 1.9909 0.13859 0 0.46038 0 0.38498
F5 0 0.37711 0 0.25357 0 0.45344 1.0136 0.40474 0 0.32898 0.15506 0.18492 0 0.52517 0 0.43327
F6 7.1435e−11 0.88702 1.3959e−05 0.25567 4.7116e−32 1.486 9.3882e−05 0.90122 4.7202e−08 0.73673 6.992e−13 0.61882 4.9019e−07 1.0446 4.7358e−32 0.90217
F7 3.0753e−04 0.37903 2.0981e−3 0.19671 1.1452e−3 0.39969 2.2021e−2 0.39899 2.0363e−2 0.23916 9.8938e−4 0.09926 3.0749e−04 0.45177 1.2232e−3 0.4013
F8 3 0.30494 3 0.24963 3 0.31474 3 0.33 3 0.16243 3 0.060188 3.2 0.36315 3 0.30783
F9 − 3.2374 0.42087 − 3.0711 0.39666 − 2.993 0.41104 − 1.9379 0.38357 − 3.203 0.25917 − 3.2031 0.12689 − 3.3136 0.45415 − 3.322 0.36787
F10 − 10.5364 0.58074 − 4.0465 0.29268 − 10.5363 0.69076 − 2.2012 0.5037 − 10.5363 0.35811 − 10.5364 0.26637 − 10.5364 0.59392 − 5.1756 0.50704
ASR 1 6 2 8 4 7 5 3
ATR​ 4 2 7 5 3 1 8 6

Table 4.  Comparison results of IBFO-A, BFO, OOA, GA, GWO, PSO, BWO, and DBO on the 10 benchmark functions.

15

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 5.  Comparison of convergence curves of 8 optimization algorithms.

solution space more extensively in the early stage of iteration, which increases the possibility of quickly search-
ing for the global optimal solution. In addition, the improved adaptive activity solves the problem that excessive
elimination-dispersal probability in the late iteration will lead to frequent update of bacterial colony location,
which makes it difficult to conduct fine search near the optimal solution. In addition, it can be found that IBFO-A
algorithm can also search the global optimal solution when Ped parameter value is low, but it needs more iterations
and time cost. From the sensitivity analysis of IBFO-A to hyperparameter pc , it can be seen that the higher the
probability of pc , the more the coding composition of the individual in the flora is affected by other individuals,

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 16

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 6.  Sensitivity analysis of IBFO-A to parameter N.

Figure 7.  Sensitivity analysis of IBFO-A to parameter Ped.

Figure 8.  Sensitivity analysis of IBFO-A to parameter pc.

F1 F7 F9
Score Time Score Time Score Time
N
30 0 0.26881 3.2398e−4 0.34955 − 3.2291 0.39857
50 0 0.28275 3.0818e−4 0.34564 − 3.2296 0.40963
80 0 0.33146 3.1231e−4 0.37591 − 3.3105 0.47671
100 0 0.35398 3.0753e−4 0.37827 − 3.3136 0.51034
Ped
0.4 0 0.43908 3.0939e−4 0.43906 − 2.4921 0.40801
0.6 0 0.4155 3.0869e−4 0.41504 − 2.6352 0.37525
0.8 0 0.34138 3.0749e−4 0.39222 − 3.2036 0.39457
0.95 0 0.36583 3.0749e−4 0.39482 − 3.3043 0.39887
pc
0.1 0 0.33018 3.0956e−4 0.32449 − 3.2927 0.40793
0.3 0 0.34028 3.1217e−4 0.34316 − 3.3085 0.40045
0.5 0 0.34936 3.0761e−4 0.34173 − 3.3193 0.43215
0.8 0 0.35788 3.0766e−4 0.39245 − 3.3164 0.42945

Table 5.  Parameter sensitivity comparison experiment.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 17

Vol.:(0123456789)
 www.nature.com/scientificreports/

thus improving the optimization ability of the flora. However, even if IBFO-A uses the improved reproductive
activity, too high a pc probability may still lead to a decrease in bacterial diversity, putting the algorithm at risk
of falling into a local optimal solution. Finally, according to the optimization results and running time analysis,
IBFO-A algorithm is not sensitive to the change of hyperparameters within a reasonable range.

Comparison and analysis of twelve optimization algorithms in CEC2019 benchmark functions

The experimental design of algorithm optimization performance comparison is as follows: In addition to the
eight optimization algorithms mentioned above, including IBFO-A, Two novel and improved optimization
algorithms for PSO and BFO :AWPSO25 and ­ChaoticBFO26, as well as the two latest advanced optimization
algorithms ­COA63 and G ­ O64, a total of 12 optimization algorithms were compared in 10 sets of different types
of advanced benchmark functions in CEC2019. We set experimental parameters uniformly for all optimization
algorithms. The specific 10 groups of optimization convergence curves, scores and running schedules are shown
in Fig. 9 and Table 6.
According to the experimental data in convergence diagram 9 and Table 6, compared with other 11 algo-
rithms, IBFO-A has the best performance in four function scoring tests, and ASR ranks first, indicating that
IBFO-A has good optimization performance. IBFO-A performs well in F1 and F10 test functions designed for
single-objective real parameter optimization, demonstrating the IBFO-A algorithm’s good performance in the
global search for the best solution. It also performs well in the two high-dimensional test functions F2 and F3,
This shows that the improved chemotactic activity and replication activity achieve a harmonious equilibrium
between exploration and exploitation. It makes IBFO-A algorithm have better searching ability in test functions
of different dimensions, and can be used to optimize DBN structure model. In addition, the basic algorithm
BFO and another improved algorithm, ChaoticBFO, are also competitive in F1 and F2 compared with other
test functions, but their performance is slightly inferior in F3 test functions, which may be because the random
elimination-dispersal activity they use is difficult to escape the local optimal solution under high-dimensional
functions. For the test function of fixed-dimensional multi-modal and multi-objective optimization, IBFO-A
ranks among the best in F7-F9 and performs well in F4-F6, indicating that IBFO-A algorithm can be applied
to multimodal and multi-objective optimization problems. In addition, the ATR of IBFO-A algorithm ranks
9th, indicating that the running time of IBFO-A algorithm has increased in complex optimization problems.

Two real‑world engineering optimization problems

In this section, we use two different real-world engineering optimization problems to evaluate the model opti-
mization capabilities of the IBFO-A algorithm, where each set of algorithms is run independently 50 times. Two
kinds of engineering optimization problem parameter selection are shown in Table 7.
The first engineering optimization problem we chose was: Tension/compression spring design problem
(TCSD)65, TCSD is a continuously constrained problem such that the volume V of the coil spring is minimized
under constant tension/compression load. The second engineering optimization problem we selected is Con-
strained truss optimization ­problem66. Three-bar truss is a common structural form in engineering, which is
widely used in bridges, buildings, mechanical equipment and other fields. The optimization problem of structure
design of three-bar truss is to get the best structure layout under certain constraints by adjusting the parameters
such as the size, shape and connection mode of the bar. The running results of 12 optimization algorithms in
TCSD and Three-bar truss engineering problems are shown in Tables 8 and 9
Two groups of experiments show that IBFO-A algorithm has improved optimization performance compared
with the original BFO algorithm in finding the best objective function. In summary, IBFO-A algorithm has a
good optimization ability in practical engineering applications.

Experimental results and analysis of algorithm convergence

The performance test experiment of the IBFO-D algorithm on network learning is conducted in two steps. In
the first step, two types of data were randomly extracted from the alarm benchmark network:

(1) Three sets of non-temporal data samples, each containing 1000, 2000, and 3000 randomly selected sample
points.
(2) Three sets of time series sample data. The time series data contains two time slices, and each set contains
1000, 2000, and 3000 randomly selected sample points.

As the optimization process is a random search, each iteration experiment is independently run 50 times to
comprehensively evaluate and analyze the iterative convergence of the fitness values of the IBFO-D algorithm
with respect to B0 and B→. This analysis checks the stability of the learning network of the algorithm and whether
it falls into its local optimal solution.
We chose to conduct convergence analysis experiments in the alarm network for two reasons:

(1) In the field of learning Bayesian network structures, the alarm network is widely recognized as the most
popular benchmark.
(2) Compared with other network structures, the alarm network is more complex, and its performance on
complex networks can better reflect the global search capability and stability of the IBFO-D algorithm.

Figures 10 and 11 illustrate the convergence of fitness scores during iterations, with the X-axis and Y-axis
representing the number of iterations and the fitness score, namely the K2 score, respectively. In each generation,

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 18

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 9.  Comparison of convergence curves of 12 optimization algorithms.

the K2 score represents the average outcome of 50 independent runs of the algorithm. An analysis of the experi-
mental findings shows that the algorithm reaches convergence at approximately 110 iterations for the three sets
of temporal data, whereas for the other three sets of non-temporal data, convergence is achieved at approximately
70 iterations. This indicates that the IBFO-D algorithm can converge stably within a high fitness value in the
temporal and non-temporal data without getting trapped in local optima due to its improved chemotaxis, repro-
duction, and elimination-dispersal strategies. Furthermore, this algorithm demonstrated a rapid convergence
speed and good convergence accuracy.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 19

Vol.:(0123456789)
Vol:.(1234567890)
Scientific Reports |
(2024) 14:8266 |
www.nature.com/scientificreports/

IBFO-A BFO OOA GA GWO PSO

Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​
F1 1 0.51471 1.1779e+8 0.13126 1 0.66938 1.7816e+8 0.57757 1.3322e+4 0.61903 6.5303e+7 1.0164
F2 4.9857 0.43262 1.0458e+4 0.20157 5 0.43205 5.6961e+3 0.51007 1.0655e+2 0.41483 1.9105e+4 0.59639
F3 2.4498 0.31644 10.837 0.085054 6.4824 0.4427 9.6893 0.25438 2.0939 0.33925 7.3875 0.47824
F4 69.4439 0.37462 78.9059 0.12656 89.3174 0.52289 20.7976 0.26875 9.0293 0.30214 34.8285 0.52836
F5 13.8409 0.35037 79.3276 0.16411 81.1214 0.55487 1.9168 0.49689 1.7881 0.35106 1.0714 0.49645
1 9 12 1 10 11 7 8 3 6 5 12
F6 7.6439 3.7762 11.216 0.72694 9.4888 4.6178 2.046 2.4695 2.2431 2.3497 4.0063 4.6507
F7 1.0269e+3 0.41371 1.8581e+3 0.15723 1.7834e+3 0.55431 7.3754e+2 0.37282 4.8324e+2 0.58203 6.0044e+2 0.80284
F8 3.8167 0.33895 4.8899 0.088168 4.7926 0.3799 4.9967 0.27517 4.4762 0.24203 3.6252 0.42283
F9 1.0912 0.36772 3.6541 0.081567 2.6682 0.4817 1.2533 0.25649 1.1158 0.22874 1.2037 0.35686

https://doi.org/10.1038/s41598-024-58806-0
F10 20.9891 0.33267 21.555 0.16646 21.3348 0.46354 21.4439 0.56025 21.535 0.25705 21.3227 0.40833
BWO DBO COA GO AWPSO ChaoticBFO
Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​ Score Time ASR|ATR​
F1 1 0.48294 1 0.45176 1 0.43022 5.9018e+8 0.4726 2.31e+9 0.36643 3.6731e+3 0.17772
F2 5 0.62992 4.2723e+2 0.44874 5 0.42175 4.998 0.60817 1.1964e+4 0.27733 73.933 0.17162
F3 3.3701 0.40246 3.1877 0.27811 4.213 0.2029 9.7112 0.28002 1.4091 0.19957 6.7682 0.1404
F4 21.5619 0.38112 23.3781 0.29205 37.838 0.37738 87.5607 0.28915 78.6063 0.4423 87.3588 0.21643
F5 2.4201 0.42112 1.1945 0.30336 1.1084 0.33756 1.6568 0.34559 8.2245 0.20874 48.5569 0.13086
4 10 6 7 2 4 9 5 8 3 11 2
F6 4.1681 2.9611 6.6196 2.4659 2.2831 3.8721 12.0047 2.4237 4.4743 2.3223 10.5205 0.76085
F7 8.1706e+2 0.63241 1.182e+3 0.41399 1.095e+3 0.3335 2.2759e+3 0.3119 1.5242e+3 0.23214 1.1435e+3 0.13789
F8 3.8176 0.50389 4.0369 0.36719 3.7907 0.33956 4.7452 0.35088 3.6329 0.21435 4.8279 0.17259
F9 1.405 0.44 1.2145 0.44716 1.3554 0.4415 1.7225 0.28468 1.4362 0.19956 3.7324 0.19608
F10 21.236 0.37332 21.2753 0.30392 20.999 0.25123 21.0001 0.29586 21.2158 0.21668 21.6367 0.13173

Table 6.  Comparison results of IBFO-A, BFO, OOA, GA, GWO, PSO, BWO, DBO, COA, GO, AWPSO, and ChaoticBFO on the CEC2019 benchmark functions.

20
www.nature.com/scientificreports/

Test function Subject to S

x23 x3
q1 (x) = 1 − 71785x14
≤0

4x22 −x1 x2 1
2 ≤ x1 ≤ 15
min f (x) = (x3 + 2)x2 x12 q2 (x) = 1256 x2 x31 −x41
  + 5108x21
−1≤0 0.25 ≤ x2 ≤ 1.3
140.45x1
0.05 ≤ x3 ≤ 2
q3 (x) = 1 − x22 x3
≤0
x2 +x1
q4 (x) = 1.5 − 1 ≤ 0
√
g1 (x) = √ 2x1 +x2  P − σ ≤0
2x12 +2x1 x2
 √ 
min f (x) = 2 2x1 + x2 × l g2 (x) = √ 2 x2 P−σ ≤0 0 ≤ xi ≤ 1, i = 1, 2
2x1 +2x1 x2

g3 (x) = √ 1 P−σ ≤0
2x2 +x1

Table 7.  Two kinds of engineering optimization problem parameter selection.

IBFO-A BFO OOA GA GWO PSO

Best 0.012667247 0.013558966 0.012668072 NAN 0.01266895 0.012666835
Worst 0.014586288 0.029029047 0.016834462 NAN 0.012729057 0.016726958
Average 0.013056487 0.021083917 0.013893121 NAN 0.012705313 0.013185834
Std 4.60e−04 3.59e−03 1.10e−03 NAN 1.95e−05 9.54e−04
BWO DBO COA GO AWPSO ChaoticBFO
Best 0.01273713 0.012674165 0.01267023 0.012669091 0.012666912 0.013769665
Worst 0.016191356 0.017773158 0.01351603 0.017044529 0.014294677 0.023413664
Average 0.013232493 0.013493124 0.012866426 0.013174405 0.012949943 0.013940312
Std 7.39e−04 1.61e−03 1.93e−04 9.22e−04 3.30e−04 4.64e−03

Table 8.  The running results of 12 optimization algorithms in TCSD engineering problems.

IBFO-A BFO OOA GA GWO PSO

Best 263.8959581 263.9397436 263.9019434 NAN 263.8959079 NAN
Worst 264.0877488 264.8430515 270.2676865 NAN 263.9024999 NAN
Average 263.9072699 264.28099 264.9638793 NAN 263.8977088 NAN
Std 3.01e−02 2.63e−01 1.30e+00 NAN 1.53e−03 NAN
BWO DBO COA GO AWPSO ChaoticBFO
Best 263.9797514 263.8958434 263.895859 263.8958465 263.8958434 263.929491
Worst 264.9257365 263.8969615 263.8975473 263.8966012 263.8963019 265.1599811
Average 264.2141197 263.8959623 263.8962099 263.8959696 263.8958947 264.1313994
Std 1.81e−01 2.00e−04 3.29e−04 1.29e−04 8.55e−05 2.58e−01

Table 9.  The running results of 12 optimization algorithms in Three-bar truss engineering problem.

Experimental results and analysis of the algorithm performance comparison

The second step involved the use of two dynamic benchmark networks as experimental models. The structural
hamming distance (SHD)67 was used to compare the IBFO-D algorithm with the DMMHC a­ lgorithm9 and the
GS ­algorithm10 based on temporal information as a comprehensive evaluation metric. This comparison was
conducted using three different sets of data samples. Each experiment was independently conducted 50 times
to ensure thorough validation of the accuracy and efficiency of these algorithms.
The performance comparison results of the three different algorithms in the small-scale 2T-Asia network and
the large-scale 2T-alarm network, across six different data samples, are presented in Tables 10, 11, 12, 13, 14 and
15. In these tables, μ ± σ denotes the average value μ and standard deviation σ of the execution time (seconds)
over 50 independent runs for each algorithm.
For the 2T-Asia network, the IBFO-D and DMMHC algorithms have the same optimal SHD when the sample
size is 1000. However, the IBFO-D algorithm exhibits better stability and accuracy than the DMMHC algorithm
with respect to the worst and average results. This is because the IBFO-D algorithm is based on global-search-
based SI optimization, where the error of an individual agent does not affect the optimization outcome of the
entire swarm. Furthermore, the reproductive activity in IBFO-D improves the information exchange capability
among bacterial individuals, thereby enhancing the overall optimization performance of the bacterial population.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 21

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 10.  Experimental results of iterative convergence for three sets of non-temporal data samples.

Figure 11.  Experimental results of iterative convergence for three sets of 2T-BN temporal data samples.

(1) From the viewpoint of structural metrics, the differences in algorithm performance become more evident as
the sample size increases and network complexity improves. The IBFO-D algorithm exhibits a clear advantage
when the sample size is 3000, with a stable SHD of 1. Compared with IBFO-D, the DMMHC algorithm may have
lower accuracy, but it generates networks that are relatively close to the true structure. However, the GS algorithm
performs the worst, exhibiting the maximum structural variation in all scenarios. (2) The DMMHC algorithm
is the fastest, closely followed by the IBFO-D algorithm, while the GS algorithm is the slowest when consider-
ing time metrics. These results can be attributed to the fact that global search typically requires more time than
greedy local search. Furthermore, in SI algorithms, optimization and complete information exchange tasks are
independently executed by individual agents during each iteration, leading to higher time costs. Notably, in the
IBFO-D algorithm, significant time savings are achieved by omitting the grouping mechanism when searching
for DBN structures in networks such as the small-scale Asia network.
For the 2T-alarm network: (1) When structural metrics are considered, the SHD values for all three algorithms
are relatively large in the dataset with a sample size of 2000. This is because a small number of sample cases may

DMMHC GS IBFO-D
SHD TIME SHD TIME SHD TIME
Best 2 41.9 8 127.2 2 44.7
Worst 4 63.2 10 191.6 3 61.1
Average 3.5 49.6 ± 11.7 9.3 144.7 ± 15.9 2.2 53.4 ± 6.3

Table 10.  Experimental results of the performance comparison of the three algorithms on 2T-Asia-1000.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 22

Vol:.(1234567890)
www.nature.com/scientificreports/

DMMHC GS IBFO-D
SHD TIME SHD TIME SHD TIME
Best 2 89.8 7 210.7 1 97.6
Worst 3 117.2 10 284.2 2 114.2
Average 2.7 101.3 ± 15.8 7.9 253.3 ± 41.3 1.6 107.4 ± 8.4

Table 11.  Experimental results of the performance comparison of the three algorithms on 2T-Asia-2000.

DMMHC GS IBFO-D
SHD TIME SHD TIME SHD TIME
Best 2 182.6 4 392.3 1 198.7
Worst 3 220.3 7 492.6 1 211.8
Average 2.2 196.3 ± 13.3 6.3 441.5 ± 47.1 1 205.2 ± 7.6

Table 12.  Experimental results of the performance comparison of the three algorithms on 2T-Asia-3000.

DMMHC GS IBFO-D
SHD TIME SHD TIME SHD TIME
Best 31 523.7 42 712.6 25 645.5
Worst 40 1006.4 61 1326.2 29 752.7
Average 34.7 783.3 ± 256.7 53.3 1026.4 ± 302.4 27.4 697.1 ± 41.4

Table 13.  Experimental results of the performance comparison of the three algorithms on 2T-alarm-2000.

DMMHC GS IBFO-D
SHD TIME SHD TIME SHD TIME
Best 24 1546.5 42 2164.1 13 1449.6
Worst 33 2152.8 58 3321.8 20 1813.1
Average 27.6 1798.8 ± 395.3 47.6 2713.5 ± 523.9 18.2 1590.1 ± 184.2

Table 14.  Experimental results of the performance comparison of the three algorithms on 2T-alarm-5000.

DMMHC GS IBFO-D
SHD TIME SHD TIME SHD TIME
Best 20 2603.6 39 3765.3 9 2414.8
Worst 26 3319.4 45 5143.8 18 2926.7
Average 22.5 3120.5 ± 368.2 42.5 4246.1 ± 663.5 12.3 2753.7 ± 226.1

Table 15.  Experimental results of the performance comparison of three algorithms on 2T-alarm-8000.

not fully reflect the network characteristics in complex networks, leading to challenges in the accurate learning
of network structure by the algorithms. However, in all scenarios, the IBFO-D algorithm consistently exhibits
smaller structural variations than the other two algorithms, with a worst-case SHD of 29. In the dataset with a
sample size of 5000, a notable enhancement in algorithm performance was observed. On average, 91.8 of 110
edges were correctly identified by the IBFO-D algorithm, making it the best-performing algorithm among the
three. In the dataset with a sample size of 8000, this advantage becomes even more pronounced. This enhance-
ment is due to the improved chemotaxis and elimination-dispersal approach, which improves the global opti-
mization capability of the IBFO-D algorithm and enables the escape from local optima, thereby facilitating the
search for the global optimum structure. (2) Considering time metrics, learning networks in complex node
sequences requires more time. Comparative experimental analysis revealed that the IBFO-D and DMMHC algo-
rithms exhibit similar execution efficiencies on large-sample datasets, indicating that the improved chemotactic
activity in the IBFO-D algorithm facilitates fast optimization for edge orientation, resulting in optimal time

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 23

Vol.:(0123456789)
 www.nature.com/scientificreports/

performance. There is one exception, where, in the 2T-alarm-2000 dataset, the DMMHC algorithm outperforms
the IBFO-D algorithm in terms of the best runtime. This is due to the lack of a local optima escape mechanism
in the DMMHC algorithm, resulting in it being trapped in local optima in complex networks with inadequate
sample size. Among the three algorithms, the GS algorithm performs the worst, mainly because of the substantial
amount of time spent searching the search space. Based on the aforementioned experimental analysis, it can be
concluded that the IBFO-D algorithm is an effective approach for learning DBNs from data, as it can identify net-
work structures with high scores and low structural variations.At the same time, it has high execution efficiency.

Conclusion
In this study, an IBFO-A was proposed using the logistics-sine chaotic mapping method to initialize the popula-
tion and improve the chemotactic activity, reproductive activity, and elimination-dispersal activity of the bacteria
by combining the OOA algorithm development stage, GA crossover idea, and adaptive method. To solve the
problem of complex DBN learning structures due to the introduction of time information, an IBFO-D algo-
rithm is proposed within the framework of the IBFO-A algorithm. In this algorithm, the fitness function and
V-structure orientation rule were constructed, and simulation experiments were conducted on a series of refer-
ence functions, the 2T-Asia network and the 2T-Alarm network. The experimental results show that the initial
population of the IBFOA algorithm using the chaotic mapping method can accelerate the iterative convergence
speed, and the improved chemotactic activity and reproductive activity can improve the optimization ability
of bacteria. Based on the adaptive elimination-dispersal activity, the algorithm can effectively prevent the local
optimal to guide bacteria to find a better solution. The IBFO-D algorithm demonstrates stable convergence at
higher fitness values in temporal and non-temporal data, and its performance is better than that of the other two
algorithms. Future work will focus on applying the IBFO-D algorithm to learn higher-order dynamic Bayesian
networks and time-varying dynamic Bayesian networks to reduce the complexity of their time computation. In
addition, the improved BFO method will be combined with other meta-heuristic methods to further improve
its ability to search for optimal datasets.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding
author on reasonable request.

Received: 17 January 2024; Accepted: 3 April 2024

References
1. Kitson, N. K., Constantinou, A. C., Guo, Z., Liu, Y. & Chobtham, K. A survey of Bayesian Network structure learning. Artif. Intell.
Rev. 56, 1–94 (2023).
2. Shiguihara, P., Lopes, A. D. A. & Mauricio, D. Dynamic Bayesian network modeling, learning, and inference: A survey. IEEE Access
9, 117639–117648 (2021).
3. Adabor, E. S. & Acquaah-Mensah, G. K. Restricted-derestricted dynamic Bayesian Network inference of transcriptional regulatory
relationships among genes in cancer. Comput. Biol. Chem. 79, 155–164 (2019).
4. Dang, S., Chaudhury, S., Lall, B. & Roy, P. K. The dynamic programming high-order dynamic Bayesian networks learning for
identifying effective connectivity in human brain from fMRI. J. Neurosci. Methods 285, 33–44 (2017).
5. Komurlu, C. et al. Active inference for dynamic Bayesian networks with an application to tissue engineering. Knowl. Inf. Syst. 50,
917–943 (2017).
6. Qu, L. et al. Dynamic Bayesian network modeling based on structure prediction for gene regulatory network. IEEE Access 9,
123616–123634 (2021).
7. Ghahramani, Z. Learning dynamic Bayesian networks. International School on Neural Networks, Initiated by IIASS and EMFCSC,
168–197 (1997).
8. Pérez-Ariza, C. B., Nicholson, A. E., Korb, K. B., Mascaro, S. & Hu, C. H. in AI 2012: Advances in Artificial Intelligence: 25th Aus‑
tralasian Joint Conference, Sydney, Australia, December 4–7, 2012. Proceedings 25. 902–913 (2012).
9. Trabelsi, G., Leray, P., Ben Ayed, M. & Alimi, A. M. in Advances in Intelligent Data Analysis XII: 12th International Symposium,
IDA 2013, London, UK, October 17–19, 2013. Proceedings 12. 392–403 (2013).
10. Trabelsi, G. New structure learning algorithms and evaluation methods for large dynamic Bayesian networks, Université de Nantes;
Ecole Nationale d’Ingénieurs de Sfax, (2013).
11. Gámez, J. A., Mateo, J. L. & Puerta, J. M. Learning Bayesian networks by hill climbing: Efficient methods based on progressive
restriction of the neighborhood. Data Min. Knowl. Disc. 22, 106–148 (2011).
12. Srinivas, M. & Patnaik, L. M. Genetic algorithms: A survey. Computer 27, 17–26 (1994).
13. Gámez, J. A. & Puerta, J. M. Searching for the best elimination sequence in Bayesian networks by using ant colony optimization.
Pattern Recogn. Lett. 23, 261–277 (2002).
14. Gheisari, S. & Meybodi, M. R. Bnc-pso: Structure learning of bayesian networks by particle swarm optimization. Inf. Sci. 348,
272–289 (2016).
15. Sahin, F. & Devasia, A. Distributed particle swarm optimization for structural Bayesian network learning. (2007).
16. Mirjalili, S., Mirjalili, S. M. & Lewis, A. Grey wolf optimizer. Adv. Eng. Softw. 69, 46–61 (2014).
17. Ji, J., Wei, H. & Liu, C. An artificial bee colony algorithm for learning Bayesian networks. Soft Computing 17, 983–994 (2013).
18. Yang, X.-S. & He, X. Bat algorithm: Literature review and applications. Int. J. Bio-inspired Comput. 5, 141–149 (2013).
19. Mirjalili, S. & Lewis, A. The whale optimization algorithm. Adv. Eng. Softw. 95, 51–67 (2016).
20. Fister, I., Fister, I. Jr., Yang, X.-S. & Brest, J. A comprehensive review of firefly algorithms. Swarm Evol. Comput. 13, 34–46 (2013).
21. Metzler, R., Chechkin, A. V., Gonchar, V. Y. & Klafter, J. Some fundamental aspects of Lévy flights. Chaos Solitons Fractals 34,
129–142 (2007).
22. Mahdavi, S., Rahnamayan, S. & Deb, K. Opposition based learning: A literature review. Swarm Evol. Comput. 39, 1–23 (2018).
23. Deng, W., Xu, J. & Zhao, H. An improved ant colony optimization algorithm based on hybrid strategies for scheduling problem.
IEEE access 7, 20281–20292 (2019).
24. Gao, Z.-M. & Zhao, J. An improved grey wolf optimization algorithm with variable weights. Comput. Intell. Neurosci. 2019 (2019).

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 24

Vol:.(1234567890)
www.nature.com/scientificreports/

25. Liu, W. et al. A novel sigmoid-function-based adaptive weighted particle swarm optimizer. IEEE Trans. Cybern. 51, 1085–1093
(2019).
26. Zhang, Q. et al. Chaos enhanced bacterial foraging optimization for global optimization. Ieee Access 6, 64905–64919 (2018).
27. Mou, J., Zhu, Q., Liu, Y. & Bai, Y. Multi-objective optimal thrust allocation strategy for automatic berthing of surface ships using
adaptive non-dominated sorting genetic algorithm III. Ocean Eng. 299, 117288 (2024).
28. Giri, P. K., De, S. S. & Dehuri, S. Adaptive neighbourhood for locally and globally tuned biogeography based optimization algo-
rithm. J. King Saud University-Comput. Inf. Sci. 33, 453–467 (2021).
29. Giri, P. K., De, S. S., Dehuri, S. & Cho, S. B. Biogeography based optimization for mining rules to assess credit risk. Intell. Syst. Acc.
Finance Manag. 28, 35–51 (2021).
30. Hemeida, A. et al. Implementation of nature-inspired optimization algorithms in some data mining tasks. Ain Shams Eng. J. 11,
309–318 (2020).
31. Zhu, B. et al. A critical scenario search method for intelligent vehicle testing based on the social cognitive optimization algorithm.
IEEE Trans. Intell. Transp. Syst. (2023).
32. Ashraf, N. M., Mostafa, R. R., Sakr, R. H. & Rashad, M. Optimizing hyperparameters of deep reinforcement learning for autono-
mous driving based on whale optimization algorithm. Plos one 16, e0252754 (2021).
33. Zhang, C., Zhou, L. & Li, Y. Pareto optimal reconfiguration planning and distributed parallel motion control of mobile modular
robots. IEEE Trans. Ind. Electron. (2023).
34. Khan, S. A. & Engelbrecht, A. P. A fuzzy particle swarm optimization algorithm for computer communication network topology
design. Appl. Intell. 36, 161–177 (2012).
35. Cao, B. et al. Multiobjective 3-D topology optimization of next-generation wireless data center network. IEEE Trans. Ind. Inform.
16, 3597–3605 (2019).
36. Nonut, A. et al. A small fixed-wing UAV system identification using metaheuristics. Cogent Eng. 9, 2114196 (2022).
37. Duan, H., Lei, Y., Xia, J., Deng, Y. & Shi, Y. Autonomous maneuver decision for unmanned aerial vehicle via improved pigeon-
inspired optimization. IEEE Trans. Aerosp. Electron. Syst. (2022).
38. Li, G., Gao, X. & Di, R. in 2014 IEEE/ACIS 13th International Conference on Computer and Information Science (ICIS). 245–250
(2014).
39. Xing-Chen, H., Zheng, Q., Lei, T. & Li-Ping, S. in 2007 IEEE Symposium on Artificial Life.85–91 (2007).
40. Santos, F. P. & Maciel, C. D. in 5th ISSNIP-IEEE Biosignals and Biorobotics Conference (2014): Biosignals and Robotics for Better
and Safer Living (BRC). 1–6 (2014).
41. Daqi, J. & Wang, H. in Proceedings of 2021 5th Chinese Conference on Swarm Intelligence and Cooperative Control. 1315–1325
(2021).
42. Quesada, D., Bielza, C. & Larrañaga, P. in International Conference on Hybrid Artificial Intelligence Systems. 158–171 (2021).
43. Deng, Y.-J., Liu, H.-R., Wang, H.-Y. & Liu, B. Learning Dynamic Bayesian Networks structure based on a new hybrid K2-Bat
learning algorithm. J. Chin. Inst. Eng. 44, 41–52 (2021).
44. Passino, K. M. Biomimicry of bacterial foraging for distributed optimization and control. IEEE Control Syst. Mag. 22, 52–67 (2002).
45. Guanglei, M. et al. A survey of Bayesian Network structure learning. J. Beihang Univ. https://​doi.​org/​10.​13700/j.​bh.​1001-​5965.​
2023.​0445 (2023).
46. Friedman, N., Murphy, K. & Russell, S. Learning the structure of dynamic probabilistic networks. https://​arxiv.​org/​abs/​1301.​7374
(2013).
47. Serfozo, R. Basics of Applied Stochastic Processes (Springer, 2009).
48. Varol Altay, E. & Alatas, B. Bird swarm algorithms with chaotic mapping. Artif. Intell. Rev. 53, 1373–1414 (2020).
49. Demir, F. B., Tuncer, T. & Kocamaz, A. F. A chaotic optimization method based on logistic-sine map for numerical function
optimization. Neural Comput. Appl. 32, 14227–14239 (2020).
50. Supriyono, H. & Tokhi, M. O. in 2010 2nd International Conference on Computational Intelligence, Communication Systems and
Networks. 72–77 (2010).
51. Niu, B., Wang, J. & Wang, H. Bacterial-inspired algorithms for solving constrained optimization problems. Neurocomputing 148,
54–62 (2015).
52. Naveen, S., Kumar, K. S. & Rajalakshmi, K. Distribution system reconfiguration for loss minimization using modified bacterial
foraging optimization algorithm. Int. J. Electr. Power Energy Syst. 69, 90–97 (2015).
53. Niu, B., Wang, H., Tan, L. & Li, L. in 2011 Seventh International Conference on Computational Intelligence and Security. 76–80
(2011).
54. Niu, B., Fan, Y., Xiao, H. & Xue, B. Bacterial foraging based approaches to portfolio optimization with liquidity risk. Neurocomput‑
ing 98, 90–100 (2012).
55. Chen, H., Zhu, Y., Hu, K. & Ma, L. Bacterial colony foraging algorithm: Combining chemotaxis, cell-to-cell communication, and
self-adaptive strategy. Inf. Sci. 273, 73–100 (2014).
56. Wang, L., Zhao, W., Tian, Y. & Pan, G. A bare bones bacterial foraging optimization algorithm. Cogn. Syst. Res. 52, 301–311 (2018).
57. Zhao, W., Wang, L., Tian, Y. & Pan, G. A bare bones bacterial foraging optimization algorithm. Cogn. Syst. Res. 52, 301–311 (2018).
58. Dehghani, M. & Trojovský, P. Osprey optimization algorithm: A new bio-inspired metaheuristic algorithm for solving engineering
optimization problems. Front. Mech. Eng. 8, 1126450 (2023).
59. De Jong, K. A. & Spears, W. M. A formal analysis of the role of multi-point crossover in genetic algorithms. Ann. Math. Artif. Intell.
5, 1–26 (1992).
60. Suganthan, P. N. et al. Problem definitions and evaluation criteria for the CEC 2005 special session on real-parameter optimization.
KanGAL Rep. 2005005, 2005 (2005).
61. Zhong, C., Li, G. & Meng, Z. Beluga whale optimization: A novel nature-inspired metaheuristic algorithm. Knowl.-Based Syst.
251, 109215 (2022).
62. Xue, J. & Shen, B. Dung beetle optimizer: A new meta-heuristic algorithm for global optimization. J. Supercomput. 79, 7305–7336
(2023).
63. Jia, H., Rao, H., Wen, C. & Mirjalili, S. Crayfish optimization algorithm. Artif. Intell. Rev. 56, 1919–1979 (2023).
64. Hamad, R. K. & Rashid, T. A. GOOSE algorithm: A powerful optimization tool for real-world engineering challenges and beyond.
Evol. Syst., 1–26 (2024).
65. Tzanetos, A. & Blondin, M. A qualitative systematic review of metaheuristics applied to tension/compression spring design prob-
lem: Current situation, recommendations, and research direction. Eng. Appl. Artif. Intell. 118, 105521 (2023).
66. Panagant, N., Kumar, S., Tejani, G. G., Pholdee, N. & Bureerat, S. Many-objective meta-heuristic methods for solving constrained
truss optimisation problems: A comparative analysis. MethodsX 10, 102181 (2023).
67. Trabelsi, G., Leray, P., Ayed, M. B. & Alimi, A. M. in 2013 5th International Conference on Modeling, Simulation and Applied Opti‑
mization (ICMSAO). 1–6 (2013).

Acknowledgements
The authors sincerely thank the anonymous reviewers for their valuable and constructive comments.

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 25

Vol.:(0123456789)
 www.nature.com/scientificreports/

Author contributions
T.T.L., C.G.W. and G.L.M. provided funding support and technical prestudy. G.L.M. and Z.L.C. devised novel
algorithms. Z.L.C. completed the programming of the algorithm and the manuscript. B.W. and M.Z.Z. set up
the experimental environment.

Funding
The authors would like to thank the Liaoning Revitalization Talents Program (no. XLYC2007144), Shenyang
Natural Science Foundation (No. 22-315-6-09), National Defense Basic Scientific Research program of China
(No. JH2023007) and National Natural Science Foundation of China (No. 62373261).

Competing interests
The authors declare no competing interests.

Additional information
Supplementary Information The online version contains supplementary material available at https://​doi.​org/​
10.​1038/​s41598-​024-​58806-0.
Correspondence and requests for materials should be addressed to Z.C.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

© The Author(s) 2024

Scientific Reports | (2024) 14:8266 | https://doi.org/10.1038/s41598-024-58806-0 26

Vol:.(1234567890)