In software engineering, testing has long been a research area of software maintenance. Testing is extremely expensive, and there is no guarantee that all defects will be found within a single round of testing. Therefore, fixing defects that are not discovered by a single round of testing is important for reducing the test costs. During the software maintenance process, testing is conducted within the scope of a set of test cases called a test suite. Mutation testing is a method that uses mutants to evaluate whether the test cases of the test suite are appropriate. In this paper, an approach is proposed that uses the mutants of a mutation test to identify defects that are not discovered through a single round of testing. The proposed method simultaneously applies two or more mutants to a single program to define and record the relationships between different lines of code. In turn, these relationships are examined using the defects that were discovered by a single round of testing, and possible defects are recommended from among the recorded candidates. To evaluate the proposed method, a comparative study was conducted using the fault localization method, which is commonly employed in defect prediction, as well as the Defects4J defect prediction dataset, which is widely used in software defect prediction. The results of the evaluation showed that the proposed method achieves a better performance than seven other fault localization methods (Tarantula, Ochiai, Opt2, Barinel, Dstar2, Muse, and Jaccard).
Citation: Dong-Gun Lee, Yeong-Seok Seo. Identification of propagated defects to reduce software testing cost via mutation testing[J]. Mathematical Biosciences and Engineering, 2022, 19(6): 6124-6140. doi: 10.3934/mbe.2022286
[1] | Catherine S. Jarnevich, Tracy R. Holcombe, Catherine Cullinane Thomas, Leonardo Frid, Aaryn Olsson . Simulating long-term effectiveness and efficiency of management scenarios for an invasive grass. AIMS Environmental Science, 2015, 2(2): 427-447. doi: 10.3934/environsci.2015.2.427 |
[2] | Sonia Prestamburgo, Filippo Sgroi, Adriano Venudo, Carlo Zanin . Universal design as resilient urban space plan strategy. New scenarios for environmental resources' sustainable management. AIMS Environmental Science, 2021, 8(4): 321-340. doi: 10.3934/environsci.2021021 |
[3] | Laura Romero, Modesto Pérez-Sánchez, P. Amparo López-Jiménez . Improvement of sustainability indicators when traditional water management changes: a case study in Alicante (Spain). AIMS Environmental Science, 2017, 4(3): 502-522. doi: 10.3934/environsci.2017.3.502 |
[4] | Amir Hedayati Aghmashhadi, Giuseppe T. Cirella, Samaneh Zahedi, Azadeh Kazemi . Water resource policy support system of the Caspian Basin. AIMS Environmental Science, 2019, 6(4): 242-261. doi: 10.3934/environsci.2019.4.242 |
[5] | Miguel Mora-Pérez, Ignacio Guillen-Guillamón, Petra Amparo López-Jiménez . A CFD study for evaluating the effects of natural ventilation on indoor comfort conditions. AIMS Environmental Science, 2017, 4(2): 289-309. doi: 10.3934/environsci.2017.2.289 |
[6] | Beatriz Gullón, Belén Gómez, José Luis Alonso, Remedios Yáñez . Environmentally friendly technologies for obtaining high sugars concentrations from invasive woody species. AIMS Environmental Science, 2015, 2(4): 884-898. doi: 10.3934/environsci.2015.4.884 |
[7] | Novi Sylvia, Husni Husin, Abrar Muslim, Yunardi, Aden Syahrullah, Hary Purnomo, Rozanna Dewi, Yazid Bindar . Design and performance of a cyclone separator integrated with a bottom ash bed for the removal of fine particulate matter in a palm oil mill: A simulation study. AIMS Environmental Science, 2023, 10(3): 341-355. doi: 10.3934/environsci.2023020 |
[8] | Winai Meesang, Erawan Baothong, Aphichat Srichat, Sawai Mattapha, Wiwat Kaensa, Pathomsorn Juthakanok, Wipaporn Kitisriworaphan, Kanda Saosoong . Effectiveness of the genus Riccia (Marchantiophyta: Ricciaceae) as a biofilter for particulate matter adsorption from air pollution. AIMS Environmental Science, 2023, 10(1): 157-177. doi: 10.3934/environsci.2023009 |
[9] | Martina Grifoni, Francesca Pedron, Gianniantonio Petruzzelli, Irene Rosellini, Meri Barbafieri, Elisabetta Franchi, Roberto Bagatin . Assessment of repeated harvests on mercury and arsenic phytoextraction in a multi-contaminated industrial soil. AIMS Environmental Science, 2017, 4(2): 187-205. doi: 10.3934/environsci.2017.2.187 |
[10] | Afshin Meraj, Tina Shoa, Fereshteh Sadeghi Naieni Fard, Hassan Mina . A comprehensive bi-objective optimization model to design circular supply chain networks for sustainable electric vehicle batteries. AIMS Environmental Science, 2024, 11(2): 279-303. doi: 10.3934/environsci.2024013 |
In software engineering, testing has long been a research area of software maintenance. Testing is extremely expensive, and there is no guarantee that all defects will be found within a single round of testing. Therefore, fixing defects that are not discovered by a single round of testing is important for reducing the test costs. During the software maintenance process, testing is conducted within the scope of a set of test cases called a test suite. Mutation testing is a method that uses mutants to evaluate whether the test cases of the test suite are appropriate. In this paper, an approach is proposed that uses the mutants of a mutation test to identify defects that are not discovered through a single round of testing. The proposed method simultaneously applies two or more mutants to a single program to define and record the relationships between different lines of code. In turn, these relationships are examined using the defects that were discovered by a single round of testing, and possible defects are recommended from among the recorded candidates. To evaluate the proposed method, a comparative study was conducted using the fault localization method, which is commonly employed in defect prediction, as well as the Defects4J defect prediction dataset, which is widely used in software defect prediction. The results of the evaluation showed that the proposed method achieves a better performance than seven other fault localization methods (Tarantula, Ochiai, Opt2, Barinel, Dstar2, Muse, and Jaccard).
It is now widely believed that the actual spread of many infectious diseases occurs in a diverse or dispersed population [1]. Subpopulations (or compartments) can be determined not only on the basis of disease-related factors such as mode of transmission, latent period, infectious period, and genetic susceptibility or resistance, but also on the basis of social, cultural, economic, demographic, and geographic factors [2]. In the viewpoint of epidemiology above, let S(t), I(t) and R(t) be the density of susceptible, infectious and recover individuals at time t, respectively, and we suppose that the dynamics of the disease transmission is governed by the following equations:
{dSdt=μN−g(S,I)−μS+γR,dIdt=g(S,I)−(μ+δ)I,dRdt=δI−μR−γR, | (1.1) |
where μ,δ,γ are all positive constants, μ the birth and death rate, δ the recovery rate of infectious individuals, and γ the rate of removed individuals who lose immunity and return to susceptible individuals class. Here, we assume that the disease does not have vertical transmission and is generally non–pathogenic and ignores the death induced by the disease. The infectious individuals force g(S,I) plays a key role in determining disease dynamics [3,4]. Traditionally, the density-dependent transmission (or the bilinear incidence rate, g(S,I)=βSI, and β is the proportionality constant) and the frequency-dependent transmission (or the standard incidence rate, g(S,I)=βSIS+I) are two extreme forms of disease transmission, which have been frequently used in well-known epidemic models [5,6,7]. There are several different nonlinear transmission functions proposed by researchers, to see more details, we refer to [8,9,10,11,12] and the references therein. Especially, Yuan and Li [13] studied a ratio-dependent nonlinear incident rate which takes the following form (1.2):
g(S,I)=f(IS)S=β(I/S)l1+α(I/S)hS=βSh−l+1IlSh+αIh, | (1.2) |
where the parameters l and h are positive constant, α is the parameter which measures the psychological or inhibitory effect. It is worthy to note that in the special case of α=1 and h=l=1, (1.2) becomes the well–known frequency-dependent transmission rate βSIS+I. In this case, the nonlinear incidence rate (1.2) can be seen as an extension form of the frequency-dependent transmission rate.
Before 1970s, ecological population modelers (involving epidemic models) typically used ordinary differential equations (ODE, e.g. model (1.1)), seeking equilibria and analyzing stability. These models provided important insights, such as when species can stably coexist and when susceptible and infectious densities oscillate over time [14]. The ODE models that have been described so far assume that the populations experience the same homogeneous environment. In reality, individual organisms are distributed in space and typically interact with the physical environment and other organisms in their spatial neighborhood [15]. More recently, many studies have shown that the spatial epidemic model is an appropriate tool for investigating the fundamental mechanism of complex spatiotemporal epidemic dynamics. In these studies, reaction-diffusion equations have been intensively used to describe spatiotemporal dynamics [16,17,18,19,20,21,22,23,24,25].
Many studies show that spatial heterogeneity generated by species dynamics is mathematically more interesting and also biologically more important [14]. In fact, relationships between individual-level processes and ecological dynamics often depend on population spatial structure, and epidemic dynamics can be governed by localized spatial processes of contact between susceptible and infectious individuals [27]. It has been suggested that spatial heterogeneity may address many of the deficiencies of epidemic models and play an important role in the spread of an epidemic [16,28,29,30,31,32,34,33]. Of them, Grenfell, Bjornstad and Kappey [31] showed that measles waves spreading from large cities to small towns in England and Wales are determined by the spatial hierarchy of the host population structure; Keeling et al. [32] showed that the spatial distribution of farms influences the regional variability of foot-and-mouth outbreaks in UK; Hufnagel et al. [34], Colizza et al. [33] showed that the high degree of predictability of the worldwide spread of infectious diseases is caused by the strong heterogeneity of the transport network; Merler and Ajelli [30] showed that spatial heterogeneity in population density results in a relevant delay in epidemic onset between urban and rural areas. Hence, understanding the role of the spatial heterogeneity in epidemic dynamics is challenging both theoretically and empirically.
In this paper, we mainly focus on the impact of spatial heterogeneity of the disease dynamics of an SIRS epidemic model corresponding to model (1.1). To incorporate the random diffusion, described by Laplacian operator [26], coming from the random wandering of susceptible, infectious and recover individuals, we add the diffusion terms to model (1.1) for the susceptible, infectious and recover individuals, respectively. To incorporate the spatial heterogeneity, we consider β:=β(x), and β(x) is a positive Hölder continuous function on ˉΩ, which is the space–dependent rate of disease transmission by infectious individuals at position x∈Ω, β(x)Il measures the infection force of the disease, and 1/(1+αIh) describes the psychological or inhibitory effect from the behavioral change of the susceptible individuals when the number of infective individuals is very large. For the sake of convenient analysis, we adopt l=1. Then the reaction-diffusion model corresponding to model (1.1) is the following model system which governs the spatial heterogeneity and population mobility:
{∂S∂t=dSΔS+μN−β(x)ShISh+αIh−μS+γR,x∈Ω,t>0,∂I∂t=dIΔI+β(x)ShISh+αIh−(μ+δ)I,x∈Ω,t>0,∂R∂t=dRΔR+δI−μR−γR,x∈Ω,t>0,∂S∂n=∂I∂n=∂R∂n=0,x∈∂Ω,t>0, | (1.3) |
where the habitat Ω⊂Rm(m≥1) is a bounded domain with smooth boundary ∂Ω (when m>1), and n is the outward unit normal vector on ∂Ω. Moreover N=S+I+R is the total population, dS, dI and dR are diffusion coefficients for the susceptible, infectious and recover individuals, respectively.
Assume that the initial values satisfy
(H1) S(x,0),I(x,0) and R(x,0) are nonnegative continuous functions in ˉΩ, ∫ΩI(x,0)dx>0 and
∫Ω(S(x,0)+I(x,0)+R(x,0))dx=N0>0. |
Mathematical models later confirm that spatial subdivision is important for the persistence of populations.
The rest of the paper is organized as follows. In Section 2, we give the global existence and uniform boundedness of solution. In Section 3, we investigate the threshold dynamics in terms of the basic reproduction number and study the asymptotic behavior of endemic equilibrium with respect to small diffusion rate of susceptible individuals. Finally, in Section 4, we provide the summary of the main results.
The first goal of this paper is to concern with the global existence of classical solutions to model (1.3).
Theorem 2.1. Model (1.3) has a unique global classical solution (S(x,t),I(x,t),R(x,t))∈[C([0,∞)×ˉΩ)∩C2,1((0,∞)×ˉΩ)]3 satisfying S(x,t),I(x,t),R(x,t)≥0 for all t>0 and
‖S(⋅,t)‖L∞(Ω)+‖I(⋅,t)‖L∞(Ω)+‖R(⋅,t)‖L∞(Ω)≤C(N0), | (2.1) |
where C(N0)>0 is a constant dependent of N0.
Proof. The local existence and uniqueness of the solutions of model (1.3) follow from a classical result ([35, Theorem 3.3.3]). It follows from the strong maximum principle [36] that S(x,t),I(x,t) and R(x,t) are nonnegative for x∈ˉΩ and t∈(0,Tmax), here Tmax is the maximal existence time for solutions of model (1.3). In what follows, we prove that the local solution can be extended to a global one, that is Tmax=∞. The method of the proof is similar to ([37,Theorem 2.2]).
Let
V(t):=∫Ω(S(x,t)+I(x,t)+R(x,t))dx |
be the total population size at time t. We can obtain that
∂∂t∫Ω(S(x,t)+I(x,t)+R(x,t))dx=∫ΩΔ(dSS+dII+dRR)dx=0,t>0. |
The population size V is a constant, i.e.,
∫Ω(S(x,t)+I(x,t)+R(x,t))dx=N0,t≥0, | (2.2) |
which shows that ‖S(⋅,t)‖L1(Ω),‖I(⋅,t)‖L1(Ω) and ‖R(⋅,t)‖L1(Ω) are bounded in [0,Tmax).
From model (1.3) we easily deduce that
{∂S∂t≤dSΔS+μI+(μ+γ)R,x∈Ω,t>0,∂I∂t≤dIΔI+(β(x)−μ−δ)I,x∈Ω,t>0,∂R∂t=dRΔR+δI−μR−γR,x∈Ω,t>0,∂S∂n=∂I∂n=∂R∂n=0,x∈∂Ω,t>0, | (2.3) |
It follows from [38, Lemma 2.1] with q=p0=1 that ‖S(⋅,t)‖L∞(Ω),‖I(⋅,t)‖L∞(Ω) and ‖R(⋅,t)‖L∞(Ω) are also bounded in [0,∞). Thus, we obtain (2.1) and complete the proof.
As expected, the steady state will play a central role in the dynamics of model (1.3). A steady state solution of model (1.3) is a time–independent (classical) solution and therefore can be viewed as a function (˜S,˜I,˜R)∈[C2(Ω)∩C1(ˉΩ)]3 satisfying
{dSΔ˜S+μ˜N−β(x)˜Sh˜I˜Sh+α˜Ih−μ˜S+γ˜R=0,x∈Ω,dIΔ˜I+β(x)˜Sh˜I˜Sh+α˜Ih−(μ+δ)˜I=0,x∈Ω,dRΔ˜R+δ˜I−(μ+γ)˜R=0,x∈Ω,∂˜S∂n=∂˜I∂n=∂˜R∂n=0,x∈∂Ω, | (2.4) |
where ˜S(x),˜I(x) and ˜R(x) denote the density of the steady state solution at location x∈Ω. In the view of (2.2), the steady state solutions also satisfy
∫Ω(˜S(x)+˜I(x)+˜R(x))dx=N0. | (2.5) |
In epidemiological model, there are two typical constant steady state solutions, namely, disease–free equilibrium and endemic equilibrium. A disease–free equilibrium (DFE) is a steady state solution of (2.4) in which ˜S(x)>0, both ˜I(x) and ˜R(x) vanish at x∈Ω, i.e., ˜I(x)=˜R(x)=0. An endemic equilibrium (EE) is a steady state solution in which ˜S(x),˜I(x),˜R(x)>0 for some x∈Ω.
This section aims to establish the threshold dynamics of (1.3) in terms of the basic reproduction number.
The basic reproductive number, denoted by R0, which is defined as the average number of secondary infections generated by a single infected individual introduced into a completely susceptible population, is one of the important quantities in epidemiology [41,44].
The DFE of model (1.3) is E0=(N0|Ω|,0,0), where |Ω| is the Lebesgue measure of Ω. Linearizing model (1.3) at E0=(N0|Ω|,0,0), we get the following system for the infection related variable I:
{∂I∂t=dIΔI+(β−(μ+δ))I,x∈Ω,t>0∂I∂n=0,x∈∂Ω,t>0. |
Using the next generation approach for spatial heterogenous populations [44], we characterize the basic reproduction number R0 for model (1.3) is
R0=1λ0, |
where λ0 is a unique positive eigenvalue with a positive eigenfunction Ψ(x) on Ω for the elliptic eigenvalues problem
{−dIΔψ+(μ+δ)ψ=λβψ,x∈Ω,∂ψ∂n=0,x∈∂Ω. | (3.1) |
Let λ∗ be the principal eigenvaluewith a positive eigenfunction ψ∗(x) on Ω for the following eigenvalue problem :
{dIΔψ+(β−(μ+δ))ψ+λψ=0,x∈Ω,∂ψ∂n=0,x∈∂Ω. | (3.2) |
We have the following result.
Lemma 3.1. sign(1−R0)=signλ∗.
Proof. Following [42,Lemma 2.3], we consider (3.1) and (3.2),
{dIΔΨ−(μ+δ)ψ+βR0Ψ=0,x∈Ω,dIΔψ∗+(β−(μ+δ))ψ∗+λ∗ψ∗=0,x∈Ω,∂Ψ∂n=∂ψ∗∂n=0,x∈∂Ω. | (3.3) |
We multiply the first equation in (3.3) by ψ∗ and the second equation in (3.3) by Ψ, integrate by parts on Ω, and subtract the two resulting equations,
(1−1R0)∫ΩβΨψ∗dx+λ∗∫ΩΨψ∗dx=0. |
Since β,Ψ,ψ∗ are positive, we have sign(1−R0)=signλ∗.
In this subsection, we show that the stability of the DFE is determined entirely by the magnitude of R0.
Theorem 3.2. For model (1.3), if R0<1, then (S(x,t),I(x,t),R(x,t))→(N0|Ω|,0,0) as t→∞. That is, the DFE is globally asymptotically stable.
Proof. Suppose that R0<1. By Lemma 3.1, it implies that λ∗>0. Observe from the second equality of model (1.3) that
{∂I∂t≤dIΔI+(β−(μ+δ))I,x∈Ω,t>0,∂I∂n=0,x∈∂Ω,t>0. | (3.4) |
Note that the following linear system
{∂Z∂t=dIΔZ+(β−(μ+δ))Z,x∈Ω,t>0,∂Z∂n=0,x∈∂Ω,t>0. | (3.5) |
admits a solution ae−λ∗tψ∗(x), and a is chosen so large that I(x,0)≤Z(x,0) for every x∈Ω. The comparison principle implies that I(x,t)≤ae−λ∗tψ∗(x), and it then follows that I(x,t)→0 as t→∞ for x∈ˉΩ.
As a result, the equation for R(x,t) is asymptotic to
∂R∂t=dRΔR−(μ+γ)R,x∈Ω,t>0. |
From the comparison principle, we can get that R(x,t)→0 as t→∞ for x∈ˉΩ. Similarly, we can get that S(x,t)→N0|Ω| as t→∞ for x∈ˉΩ. This yields the desired result.
Suppose that X=C(¯Ω;R3) have a supremum norm ‖⋅‖, then X is an ordered Banach space with the cone P consisting of all nonnegative functions in X, and X has nonempty interior, denoted by int(P). Set
X0={W=(S,I,R)∈X|∫Ω(S+I+R)dx=N0} |
and U=P∩X0. It is easy to verify that model (1.3) defines a dynamic system on U. Denote the unique solution of model (1.3) with initial value (S0,I0,R0)∈U by Φt(S0,I0,R0)=(S(⋅,t),I(⋅,t),R(⋅,t)) for any t>0. Φt is continuous and compact for t>0. Φt is pointwisely dissipative. Therefore, Φt has a global attractor [46].
Theorem 3.3. If R0>1, model (1.3) admits at least one endemic equilibrium.
Proof. We appeal to the uniform persistence theory developed in [46,47]. Denote
U0:={(S0,I0,R0)∈U|I0≠0},∂U0:={(S0,I0,R0)∈U|I0=0}. |
Note that U=U0∪∂U0. Moreover, U0 and ∂U0 are relatively open and closed subsets of U, respectively, and U0 is convex. We divide the proof into three steps.
Step 1. A direct result of the strong maximum principle for parabolic equations is ΦtU0⊂U0 for all t>0.
Step 2. Let A∂ be the maximal positively invariant set for Φt in ∂U0, i.e.
A∂:={W0∈U|Φt(W0)∈∂U0}. |
It is easy to verify that A∂:={W0∈U|I0=0}.
Denote ω(W0) as the ω–limit set of W0 in U and
ˆA∂:=∪{W0∈A∂}ω(W0). |
We now prove ˆA∂={E0}. For any W0∈A∂, i.e. I0=0, then I(x,t)=0 for all x∈Ω,t≥0, and model (1.3) becomes
{∂S∂t=dSΔS+(μ+γ)R,x∈Ω,t>0,∂R∂t=dRΔR−μR−γR,x∈Ω,t>0,∂S∂n=∂R∂n=0,x∈∂Ω,t>0, |
which implies R(⋅,t)→0,S(⋅,t)→N0|Ω| uniformly as t→∞. Hence, ˆA∂={E0}. Therefore, {E0} is a compact and isolated invariant set for Φt restricted in A∂.
Step 3. We prove that there exists some constant ϵ0 independent of initial values such that
limt→∞‖Φt(W0)−E0‖>ϵ0. |
Suppose, on the contrary, that for any ϵ1>0, there exists some initial value W∗0 such that
limt→∞‖Φt(W∗0)−E0‖≤ϵ12. | (3.6) |
For any given small ϵ2>0, let λ∗(ϵ2) be the unique principal eigenvalue of the following eigenvalue problem with a positive eigenfunction ϕI
{−dIΔϕI−(β(N0|Ω|−ϵ2)h(N0|Ω|−ϵ2)h+αϵh2−(μ+δ))ϕI=λ∗(ϵ2)ϕI,x∈Ω,∂ϕI∂n=0,x∈∂Ω. |
Note that limϵ2→0λ∗(ϵ2)=λ∗<0, where λ∗ is the principal eigenvalue of eigenvalue problem (3.2). Therefore, we can choose ϵ2 such that λ∗(ϵ2)<0. Since ϵ1 is arbitrary, we choose ϵ1=2ϵ2. In view of (3.6), there exists T>0 such that
N0|Ω|−ϵ2≤S∗(x,t)≤N0|Ω|+ϵ2,I∗(x,t),R∗(x,t)≤ϵ2,∀x∈¯Ω,∀t≥T. |
By the strong maximum principal of parabolic equations, (S∗(⋅,x),I∗(⋅,x),R∗(⋅,x))∈int(P) for all t>0. Then we can choose a sufficiently small number c∗>0 such that I∗(T,x)≥c∗ϕI. Note that c∗e−λ∗(ϵ2)(t−T)ϕI is a solution of the following linear system
{∂ˆI∂t=dIΔˆI+(β(N0|Ω|−ϵ2)h(N0|Ω|−ϵ2)h+αϵh2−(μ+δ))ˆI,x∈Ω,t>0,∂ˆI∂n=0,x∈∂Ω,t>0,ˆI(x,T)=c∗ϕI,x∈Ω. | (3.7) |
It follows from the comparison principal that
I∗(x,t)≥c∗e−λ∗(ϵ2)(t−T)ϕI,∀t≥T, |
and, hence, I∗(x,t)→∞ uniformly in ¯Ω as t→∞, which contradicts (3.6).
The result of Step 3 implies that {E0} is an isolated invariant set for Φt in U, and WS({E0})∩U0 is an empty set, where WS({E0}) is the stable set of {E0} for Φt.
Finally, Combining Steps 1–3 and [46,Theorem 1.3.1], we have that Φt is uniformly persistent with respect to (U,∂U0). Moreover, by [46,Theorem 1.3.7], model (1.3) admits at least one endemic equilibrium.
In this section, we are concerned with the asymptotic behavior of the EE of model (1.3). Our aim is to investigate the effect of the slow movement of susceptible individuals on the spatial distribution of the infectious disease. From now on, unless otherwise specified, we always assume R0>1 and
(H2) β(x)>μ for all x∈ˉΩ.
Consider the linear eigenvalue problem
{−dRΔψ+(μ+γ)(1−δβ−μ)ψ=λψ,x∈Ω,∂ψ∂n=0,x∈∂Ω, | (3.8) |
and denote the smallest eigenvalue of (3.8) by λ1:=λ1(−dRΔ+(μ+γ)(1−δβ−μ)).
Denote ξ:=dS˜S(x)+dI˜I(x)+dR˜R(x) and set
S(x)=˜S(x)ξ,I(x)=˜I(x)ξ,R(x)=˜R(x)ξ. |
Model (2.4) is equivalent to
{dIΔI+β(x)ShISh+αIh−(μ+δ)I=0,x∈Ω,dRΔR+δI−μR−γR=0,x∈Ω,dSS+dII+dRR=1,x∈Ω,∂S∂n=∂I∂n=∂R∂n=0,x∈∂Ω. | (3.9) |
The following results hold:
Lemma 3.4. (˜S(x),˜I(x),˜R(x)) is a solution of model (2.4) if and only if (S(x),I(x),R(x)) is a solution of model (3.9). Moreover
ξ=N0∫Ω(S+I+R)dx. |
Lemma 3.5. Assume that R0>1. For model (3.9), I→I∗,R→R∗ in C1(ˉΩ) as dS→0 for some I∗,R∗∈C1(ˉΩ) with I∗≥0,R∗>0 on ˉΩ.
Proof. In the view of dSS+dII+dRR=1, β(x)ShISh+αIh is uniformly bounded for any dS>0. It follows from Lp-estimate that ‖I‖W2,p is bounded for any p>1. Thus, ‖I‖C1,τ is bounded for any τ∈(0,1) by Sobolev embedding theorem. Passing to a subsequence if necessary, I→I∗ in C1(Ω) as dS→0 where I∗(x)≥0 for x∈Ω and ∂I∂n=0 for x∈Ω. By similar arguments, R→R∗ in C1(Ω) as dS→0 where R∗(x)≥0 for x∈Ω, which satisfies
{dRΔR∗+δI∗−(μ+γ)R∗=0,x∈Ω,∂R∗∂n=0,x∈∂Ω. | (3.10) |
Now we show that I∗(x)≢ on \Omega by contradiction argument. If I^*(x) = 0 , then we obtain by (3.10) that R^* = 0 , which implies that S\rightarrow\infty is almost everywhere (abbreviate a.e.) as d_S\rightarrow0 . Thus
\begin{equation} \dfrac{\beta(x)S^h}{S^h+\alpha I^h}\rightarrow \beta(x)\,\,\,\,\mathrm{ a.e\,\, as\,\, } d_S\rightarrow0. \end{equation} | (3.11) |
Define
M = \|I\|_{L^\infty(\Omega)}+\|R\|_{L^\infty(\Omega)},\, \hat{I} = \dfrac{I}{M},\,\hat{R} = \dfrac{R}{M}. |
Note that \hat{I}, \hat{R} > 0 and \|\hat{I}\|_{L^\infty(\Omega)}+\|\hat{R}\|_{L^\infty(\Omega)} = 1 . Then by a standard compactness argument for elliptic equations, after passing to a further subsequence if necessary,
\hat{I}\rightarrow \hat{I}^*,\,\hat{R}\rightarrow \hat{R}^*,\,\,\,x\in \bar{\Omega}, \,\,\,\,\mathrm{as}\,\,\,\,d_S\rightarrow0, |
where \hat{I}^*, \hat{R}^*\geq0 for x\in\Omega and
\begin{equation} \left\{\begin{array}{ll} \|\hat{I}^*\|_{L^\infty(\Omega)}+\|\hat{R}^*\|_{L^\infty(\Omega)} = 1,\\[1ex] \dfrac{\partial \hat{I}^*}{\partial \bf{n}} = \dfrac{\partial \hat{R}^*}{\partial \bf{n}} = 0, & x\in \partial\Omega. \end{array}\right. \end{equation} | (3.12) |
It follows from (3.11) that \hat{I}^* is a weak solution of
\begin{equation} \left\{\begin{array}{ll} d_I\Delta \hat{I}^*+(\beta(x) -(\mu+\delta))\hat{I}^* = 0, &x\in \Omega,\\[1ex] \dfrac{\partial \hat{I}^*}{\partial \bf{n}} = 0, & x\in \partial\Omega. \end{array}\right. \end{equation} | (3.13) |
By elliptic regularity, we have \hat{I}^*\in C^2({\bar{\Omega}}) , which gives
\begin{equation*} \left\{\begin{array}{ll} d_I\Delta \hat{I}^*+(\beta(x) -(\mu+\delta))\hat{I}^* = 0, &x\in \Omega,\\[1ex] d_R\Delta \hat{R}^*+\delta\hat{I}^* -(\mu+\gamma))\hat{R}^* = 0, &x\in \Omega,\\[1ex] \dfrac{\partial \hat{I}^*}{\partial \bf{n}} = \dfrac{\partial \hat{R}^*}{\partial \bf{n}} = 0, & x\in \partial\Omega. \end{array}\right. \end{equation*} |
It follows from maximum principle together with (3.12) that \hat{I}^*(x), \hat{R}^*(x) > 0 . We conclude that (\lambda, \psi) = (0, \hat{I}^*(x)) is a solution of (3.2). Since \hat{I}^*(x) > 0 on \Omega , it must be that \lambda^* = 0 , which implies that \mathcal{R}_0 = 1 . This contradiction yields I^*(x)\not\equiv0 . Therefore, again by maximum principle together with (3.10), we obtain R^* > 0 .
Note that I(x), R(x) > 0 for any x\in\Omega , d_S > 0 . Denote
M(x) = d_I I+d_R R. |
Let
\begin{array}{ll} J^+: = \{ x\in\bar{\Omega}|M^*(x) = 1\},\\[2ex] J^-: = \{x\in\bar{\Omega}|0 \lt M^*(x) \lt 1\}, \end{array} |
where M^*(x): = d_II^*+d_RR^* . Observe that J^-\bigcup J^+ = \bar{\Omega}.
Lemma 3.6. Assume that \mathcal{R}_0 > 1 .
(i) The set J^+ has positive Lebesgue measure.
(ii) If further assume that \lambda_1 < 0 , then the set J^- has positive Lebesgue measure.
Proof. We prove |J^+| > 0 by contradiction. If |J^+| = 0 , i.e., 0 < M^*(x) < 1 on \Omega a.e., then it follows from d_SS = 1-d_II -d_RR that S\rightarrow\infty a.e. as d_S\rightarrow0 and thus \dfrac{\beta(x)S^hI}{S^h+\alpha I^h}\rightarrow \beta(x)I^* a.e. as d_S\rightarrow0 . Therefore, I^* is a weak solution of
\begin{equation*} \left\{\begin{array}{ll} d_I\Delta I^*+(\beta(x) -(\mu+\delta))I^* = 0, &x\in \Omega,\\[1ex] \dfrac{\partial I^*}{\partial \bf{n}} = 0, & x\in \partial\Omega. \end{array}\right. \end{equation*} |
By elliptic regularity, we have I^*\in C^2(\bar{\Omega}) , which yields
\begin{equation} \left\{\begin{array}{ll} d_I\Delta I^*+(\beta(x) -(\mu+\delta))I^* = 0, &x\in \Omega,\\[1ex] d_R\Delta R^*+\delta I^* -(\mu+\gamma))R^* = 0, &x\in \Omega,\\[1ex] \dfrac{\partial I^*}{\partial \bf{n}} = \dfrac{\partial R^*}{\partial \bf{n}} = 0, & x\in \partial\Omega. \end{array}\right. \end{equation} | (3.14) |
In light of (3.14) and R^* > 0 , we have I^* > 0 . We conclude that (\lambda, \psi) = (0, I^*(x)) is a solution of (3.2). Since I^*(x) > 0 on \Omega , it must be that \lambda^* = 0 , which implies that \mathcal{R}_0 = 1 . This contradiction implies |J^+| > 0 .
We next prove part (ⅱ) by contradiction. Now assume that |J^-| = 0 , i.e., M^*(x) = 1 on \Omega a.e.. Denote
f(x) = \dfrac{\beta(x)S^hI}{S^h+\alpha I^h}-\mu I-(\mu+\gamma)R |
and choose \varphi\in C^1(\bar{\Omega}) such that \varphi\geq0 on \Omega . Multiplying the first two equations in (3.9) by \varphi , adding them together and integrating on \Omega , we have
\begin{equation} - \int_\Omega\nabla \varphi\cdot\nabla(d_II+d_RR){dx}+ \int_\Omega \varphi f(x)dx = 0. \end{equation} | (3.15) |
As d_S\rightarrow 0 , M(x)\rightarrow M^*(x) = 1 a.e on \Omega . Thus, we obtain
\begin{equation} \int_\Omega \varphi f(x)dx\rightarrow0 \,\,\mathrm{as}\,\,\,d_S\rightarrow0 \end{equation} | (3.16) |
for any \varphi\in C(\bar{\Omega}) such that \varphi\geq0 on \Omega .
Let \psi_0 be a positive eigenfunction of \lambda_1\left(-d_R\Delta +(\mu+\gamma)\left(1-\dfrac{\delta}{\beta-\mu}\right)\right) , i.e.
\begin{equation} \left\{\begin{array}{ll} -d_R\Delta\psi_0+(\mu+\gamma)\left(1-\dfrac{\delta}{\beta(x)-\mu}\right)\psi_0 = \lambda_1 \psi_0,&x\in\Omega,\\[1ex] \dfrac{\partial \psi_0}{\partial\bf{n}} = 0,&x\in\partial\Omega. \end{array}\right. \end{equation} | (3.17) |
Since -d_R\Delta R+(\mu+\gamma)R-\delta I = 0 and S, I, R > 0 on \Omega , we have
\begin{equation} \begin{array}{ll} -d_R\Delta R+(\mu+\gamma)\left(1-\dfrac{\delta}{\beta-\mu}\right)R = \dfrac{\delta}{\beta(x)-\mu}\left(\beta I-\mu I-(\mu+\gamma)R\right) \geq\dfrac{\delta}{\beta(x)-\mu}f(x),\quad x\in\Omega. \end{array} \end{equation} | (3.18) |
Multiplying (3.18) by \psi_0 , integrating by parts over \Omega and applying (3.17), we obtain
\lambda_1 \int_\Omega\psi_0Rdx \gt \int_\Omega\dfrac{\delta}{\beta(x)-\mu}\psi_0f(x)dx. |
Let d_S\rightarrow0 . It immediately follows from (3.16) that \lambda_1 \int_\Omega\psi_0R^*dx\geq0 . Since \psi_0, R^* > 0 on \Omega , we see that \lambda_1\geq0 . This contradiction yields (ⅲ).
Theorem 3.7. Let (H2) hold. Assume that \mathcal{R}_0 > 1 and \lambda_1\left(-d_R\Delta+(\mu+\gamma)(1-\frac{\delta}{\beta-\mu})\right) < 0 , then the following assertions hold:
(i) As d_S\rightarrow0 , \tilde{S} subject to a sequence,
\tilde{S}\rightarrow\tilde{S}^* = \dfrac{N_0(1-M^*(x))}{ \int_\Omega(1-M^*(x))dx} |
for some M^*(x) satisfying 0 < M^*(x)\leq1 in \Omega . Moreover, S^* = 0 on J^+\subset \bar{ \Omega} , S^* > 0 on J^-\subset \bar{ \Omega} and \int_\Omega \tilde{S}^* = N_0 .
(ii) There exist positive constants C_1, C_2 , independent of d_S , such that for sufficiently small d_S
C_1\leq \dfrac{\tilde{I}}{d_S},\dfrac{\tilde{R}}{d_S}\leq C_2. |
That is \tilde{I}, \tilde{R}\rightarrow0 uniformly in \Omega as d_S\rightarrow0 .
Proof. By (3.9), we have
\begin{array}{ll} N_0& = \int_\Omega (\tilde{S}+\tilde{I}+\tilde{R}){dx}\\[2ex] & = \xi \int_\Omega \left(\dfrac{1-d_II-d_RR}{d_S}+I+R\right)){dx},\\[2ex] & = \dfrac{\xi}{d_S}\left( \int_\Omega d_S(I+R){dx}+ \int_\Omega (1-M(x))dx\right). \end{array} |
It follows from S, I, R > 0 and d_SS+d_II+d_RR = 1 that I, R are uniformly bounded with respect to d_S . Thus,
\int_\Omega d_S(I+R)dx\rightarrow0\,\,\,\mathrm{as}\,\,\,d_S\rightarrow0. |
In view of Lemmas 3.5 and 3.6, we have
\begin{equation*} \int_\Omega(1-M(x))dx\rightarrow \int_\Omega(1-M^*(x))dx \gt 0 \quad\mathrm{as}\quad d_S\rightarrow0. \end{equation*} |
Therefore,
\begin{equation} \dfrac{\xi}{d_S}\rightarrow \dfrac{N_0}{ \int_\Omega(1-M^*(x))dx}\quad\mathrm{as}\quad d_S\rightarrow0. \end{equation} | (3.19) |
This limit is well-defined because J^- has positive measure. It follows from \tilde{S} = \dfrac{\xi}{d_S}(1-M(x)) that
\tilde{S}\rightarrow\tilde{S}^* = \dfrac{N_0(1-M^*(x))}{ \int_\Omega(1-M^*(x))dx} \quad \mathrm{as}\quad d_S\rightarrow0,\,\,x\in C^1(\bar{\Omega}), |
and \int_\Omega \tilde{S}^* = N_0 .
Now we verify (ii). It follows from d_SS+d_II+d_RR = 1 and \tilde{I} = \dfrac{\xi}{d_S}d_SI, \tilde{R} = \dfrac{\xi}{d_S}d_SR that
\begin{equation*} 0 \lt \dfrac{\tilde{I}}{d_S},\dfrac{\tilde{R}}{d_S} \lt \dfrac{\xi}{d_S} \max\left\{\dfrac{1}{d_I},\dfrac{1}{d_R}\right\}. \end{equation*} |
Hence (3.19) implies that
\begin{equation} \limsup\limits_{d_S\rightarrow0}\sup\dfrac{\tilde{I}}{d_S}, \,\, \limsup\limits_{d_S\rightarrow0}\sup\dfrac{\tilde{R}}{d_S}\leq \dfrac{N_0}{ \int_\Omega(1-M^*(x))dx} \max\left\{\dfrac{1}{d_I},\dfrac{1}{d_R}\right\}. \end{equation} | (3.20) |
Next, by contradiction, we prove
\begin{equation*} \label{IRmin} \min\left\{ \inf\limits_{\Omega}\tilde{I},\, \inf\limits_{\Omega}\tilde{R}\right\}/d_S \not\rightarrow0, \quad \mathrm{as}\quad d_S\rightarrow0. \end{equation*} |
Assume that \min\left\{ \inf_{\Omega}\tilde{I}, \, \inf_{\Omega}\tilde{R}\right\}/d_S = o(d_S) . By [48,Lemma 2.3] and (2.4), there exists a positive constant \theta such that
\begin{equation*} \begin{array}{ll} \inf\limits_{\Omega}\tilde{I}\geq \theta \int_\Omega\dfrac{\beta(x)\tilde{S}^h\tilde{I}}{\tilde{S}^h +\alpha\tilde{I}^h} = \theta(\mu+\delta) \int_\Omega\tilde{I}dx,\quad \inf\limits_{\Omega}\tilde{R}\geq \theta (\mu+\gamma) \int_\Omega\tilde{R}dx = \theta\delta \int_\Omega\tilde{I}dx. \end{array} \end{equation*} |
Hence
\int_\Omega\tilde{I}dx, \int_\Omega\tilde{R}dx = o(d_S), |
which implies
\begin{equation} \int_\Omega\dfrac{d_I\tilde{I}+d_R\tilde{R}}{d_S}dx\rightarrow0\quad \mathrm{as}\quad d_S\rightarrow0. \end{equation} | (3.21) |
Noting that
\begin{array}{ll} N_0 = \int_\Omega\dfrac{\xi}{d_S}dx - \int_\Omega\dfrac{d_I\tilde{I}+d_R\tilde{R}}{d_S}dx+ \int_\Omega (\tilde{I}+\tilde{R})dx, \end{array} |
and combining (3.19) and (3.21), we can obtain
\begin{array}{ll} N_0\rightarrow\dfrac{N_0|\Omega|}{ \int_\Omega(1-M^*(x))dx} \quad \mathrm{as}\quad d_S\rightarrow0, \end{array} |
which contradicts Lemma 3.6 (ⅱ) (i.e., |J^-| > 0) . We complete the proof of part (ⅱ).
It is now widely believed that the mathematical models have been revealed as a powerful tool to understand the mechanism that underlies the spread of the disease. In this paper, we proposed an epidemic model with ratio-dependent incidence rate incorporating both mobility of population and spatial heterogeneity, and focus on how spatial diffusion and environmental heterogeneity affect the basic reproductive number and disease dynamics of the model.
The value of our study lies in two aspects. Mathematically, we prove that the basic reproductive number \mathcal{R}_0 can be used to govern the threshold dynamics of the model: if \mathcal{R}_0 < 1 , the unique DFE is globally asymptotically stable (see Theorem 3.2), while if \mathcal{R}_0 > 1 , there is at least one endemic equilibrium (see Theorem 3.3). Epidemiologically, we find that restricting the movement of susceptible population can effectively control the number of infectious individuals (see Theorem 3.7 (ⅱ)). Simply speaking, our results may provide some potential applications in disease control.
The authors would like to thank the anonymous referees for very helpful suggestions and comments which led to improvement of our original manuscript. This research was supported by the National Natural Science Foundation of China (Grant numbers 61672013, 11601179 and 61772017), and the Huaian Key Laboratory for Infectious Diseases Control and Prevention (HAP201704).
The authors declare that they have no competing interests.
[1] |
R. H. Ur, R. Mushtaq, A. Palwasha, K. Mukhtaj, I. Nadeem, K. H. Ullah, Making the sourcing decision of software maintenance and information technology, IEEE Access, 9 (2021), 11492-11510. https://doi.org/10.1109/ACCESS.2021.3051023 doi: 10.1109/ACCESS.2021.3051023
![]() |
[2] |
F.S. Ana M, M. R. Chaudron, M. Genero, An industrial case study on the use of UML in software maintenance and its perceived benefits and hurdles, Empirical Software Eng., 23 (2018), 3281-3345. https://doi.org/10.1007/s10664-018-9599-4 doi: 10.1007/s10664-018-9599-4
![]() |
[3] |
E. Vahid, O. Bushehrian, G. Robles, Task assignment to counter the effect of developer turnover in software maintenance: A knowledge diffusion model, Inf. Software Technol., 143 (2022), 106786. https://doi.org/10.1016/j.infsof.2021.106786 doi: 10.1016/j.infsof.2021.106786
![]() |
[4] |
K. Jang, W. Kim, A method of activity-based software maintenance cost estimation for package software, J. Supercomput., 78 (2021), 8151-8171. https://doi.org/10.1007/s11227-020-03610-6 doi: 10.1007/s11227-020-03610-6
![]() |
[5] |
T. Masateru, M. Akito, M. Kenichi, O. Sawako, O. Tomoki, Analysis of work efficiency and quality of software maintenance using cross-company dataset, IEICE Trans. Inf. Syst., 104 (2021), 76-90. https://doi.org/10.1587/transinf.2020MPP0004 doi: 10.1587/transinf.2020MPP0004
![]() |
[6] | K.W. Kim, Y. Son, Software weakness evaluation method for secure software development, in Proceedings on 2021 International Conferences on Multimedia Information Technology and Applications, (2021), 322-325. |
[7] | C. Kim, D. Kim, H. Kang, Detecting defect in headlamp housing with machine learning techniques, in Proceedings on 2021 International Conferences on Multimedia Information Technology and Applications, (2021), 428-430. |
[8] |
Y. J. Choi, Y. W. Lee, B. G Kim, Residual-based graph convolutional network for emotion recognition in conversation for smart Internet of Things, Big Data, 9 (2021), 279-288. https://doi.org/10.1089/big.2020.0274 doi: 10.1089/big.2020.0274
![]() |
[9] |
P. P. Roy, P. Kumar, B.G. Kim, An efficient sign language recognition (SLR) system using Camshift tracker and hidden Markov model (hmm), SN Comput. Sci., 2 (2021), 1-15. https://doi.org/10.1007/s42979-021-00485-z doi: 10.1007/s42979-021-00485-z
![]() |
[10] | B. George, F. Stefan, M. Michael, P. Josef, An early investigation of unit testing practices of component-based software systems, in 2020 IEEE Workshop on Validation, Analysis and Evolution of Software Tests, (2020), 12-15. https://doi.org/10.1109/VST50071.2020.9051632 |
[11] | M. Alcon, H. Tabani, J. Abella, F. J. Cazorla, Enabling Unit Testing of Already-Integrated AI Software Systems: The Case of Apollo for Autonomous Driving, in 2021 24th Euromicro Conference on Digital System Design, (2021), 426-433. https://doi.org/10.1109/DSD53832.2021.00071 |
[12] | D. Xavier, S. Panichella, A. Gambi, Java unit testing tool competition: Eighth round, in Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops, (2020), 545-548. https://doi.org/10.1145/3387940.3392265 |
[13] | T. Mengqing, J. Yan, W. Xiang, P. Rushu, Black-box approach for software testing based on fat-property, in MATEC Web of Conferences, 309 (2020), 02008. https://doi.org/10.1051/matecconf/202030902008 |
[14] | S. Bo, Y. Shao, C. Chen, Study on the automated unit testing solution on the linux platform, in 2019 IEEE 19th International Conference on Software Quality, Reliability and Security Companion, (2019), 358-361. https://doi.org/10.1109/QRS-C.2019.00073 |
[15] |
M. Héctor D, J. Gunel, S. Federica, T. Paolo, C. David, Diversifying focused testing for unit testing, ACM Trans. Software Eng. Method., 30 (2021), 1-24. https://doi.org/10.1145/3447265 doi: 10.1145/3447265
![]() |
[16] | X. Wang, L. Wei, B. Tao, S. Ji, A study about unit testing for embedded software of control system in nuclear power plant, in International Symposium on Software Reliability, Industrial Safety, Cyber Security and Physical Protection for Nuclear Power Plant, (2020), 157-163. https://doi.org/10.1007/978-981-16-3456-7_17 |
[17] |
F. Anfal A, R. G. Alsarraj, A. M. Altaie, Software cost estimation based on dolphin algorithm, IEEE Access, 8 (2020), 75279-75287. https://doi.org/10.1109/ACCESS.2020.2988867 doi: 10.1109/ACCESS.2020.2988867
![]() |
[18] |
C. Sonia, H. Singh, Optimizing design of fuzzy model for software cost estimation using particle swarm optimization algorithm, Int. J. Comput. Intell. Appl., 19 (2020), 2050005. https://doi.org/10.1142/S1469026820500054 doi: 10.1142/S1469026820500054
![]() |
[19] |
A. Farrukh, A review of machine learning models for software cost estimation, Rev. Comput. Eng. Res., 6 (2019), 64-75. https://doi.org/10.18488/journal.76.2019.62.64.75 doi: 10.18488/journal.76.2019.62.64.75
![]() |
[20] |
K. Ishleen, N. G. Singh, W. Ritika, J. Vishal, B. Anupam, Neuro fuzzy—COCOMO Ⅱ model for software cost estimation, Int. J. Inf. Technol., 10 (2018), 181-187. https://doi.org/10.1007/s41870-018-0083-6 doi: 10.1007/s41870-018-0083-6
![]() |
[21] |
J. Miller, S. Wienke, M. Schlottke-Lakemper, M. Meinke, M. S. Müller, Applicability of the software cost model COCOMO Ⅱ to HPC projects, Int. J. Comput. Sci. Eng., 17 (2018), 283-296. https://doi.org/10.1504/IJCSE.2018.095849 doi: 10.1504/IJCSE.2018.095849
![]() |
[22] | A. Asheeri, M. Mohd, M. Hammad, Machine learning models for software cost estimation, in 2019 International Conference on Innovation and Intelligence for Informatics, Computing, and Technologies, (2019), 1-6. https://doi.org/10.1109/3ICT.2019.8910327 |
[23] | A. Shaina, N. Mishra, Software cost estimation using artificial neural network, in Soft Computing: Theories and Applications, Springer, (2018), 51-58. https://doi.org/10.1007/978-981-10-5699-4_6 |
[24] |
S. S. Pratap, V. P. Singh, A. K. Mehta, Differential evolution using homeostasis adaption based mutation operator and its application for software cost estimation, J. King Saud Univ.-Comput. Inf. Sci., 33 (2021), 740-752. https://doi.org/10.1016/j.jksuci.2018.05.009 doi: 10.1016/j.jksuci.2018.05.009
![]() |
[25] |
S. W. Ahmad, G. R. Bamnote, Whale-crow optimization (WCO)-based optimal regression model for software cost estimation, Sādhanā, 44 (2019), 1-15. https://doi.org/10.1007/s12046-019-1085-1 doi: 10.1007/s12046-019-1085-1
![]() |
[26] |
J. A. Khan, S. U. R. Khan, J. Iqbal, I. U. Rehman, Empirical investigation about the factors affecting the cost estimation in global software development context, IEEE Access, 9 (2021), 22274-22294. https://doi.org/10.1109/ACCESS.2021.3055858 doi: 10.1109/ACCESS.2021.3055858
![]() |
[27] | V. S. Desai, R. Mohanty, ANN-Cuckoo optimization technique to predict software cost estimation, in 2018 Conference on Information and Communication Technology, (2018), 1-6. https://doi.org/10.1109/INFOCOMTECH.2018.8722380 |
[28] | R. C. A. Alves, D. A. G. Oliveira, G. A. N. Segura, C. B. Margi, The cost of software-defining things: A scalability study of software-defined sensor networks, IEEE Access 7 (2019), 115093-115108. https://doi.org/10.1109/ACCESS.2019.2936127 |
[29] | D. G. Lee, Y. S. Seo, Testing cost reduction using nested mutation testing, in Proceedings on 2021 International Conferences on Multimedia Information Technology and Applications, (2021), 462-464. |
[30] |
N. Li, M. Shepperd, Y. Guo, A systematic review of unsupervised learning techniques for software defect prediction, Inf. Software Technol., 122 (2020), 106287. https://doi.org/10.1016/j.infsof.2020.106287 doi: 10.1016/j.infsof.2020.106287
![]() |
[31] | F. Keller, L. Grunske, S. Heiden, A. Filieri, A. Hoorn, D. Lo, A critical evaluation of spectrum-based fault localization techniques on a large-scale software system, in 2017 IEEE International Conference on Software Quality, Reliability and Security, (2017), 114-125. https://doi.org/10.1109/QRS.2017.22 |
[32] | The cost of poor software quality in the US: A 2020 report, Report of Proc. Consortium Inf. Softw. QualityTM, 2021. Available from: https://www.disputesoft.com/wp-content/uploads/2021/01/CPSQ-2020-Software-Report.pdf. |
[33] | 'Fully weaponized' software bug poses a threat to Minecraft gamers and apps worldwide including Google, Twitter, Netflix, Spotify, Apple iCloud, Uber and Amazon, 2021. Available from: https://www.dailymail.co.uk/news/article-10297693/Global-race-patch-critical-computer-bug.html. |
[34] |
P. Vitharana, Defect propagation at the project-level: results and a post-hoc analysis on inspection efficiency, Empirical Software Eng., 22 (2017), 57-79. https://doi.org/10.1007/s10664-015-9415-3 doi: 10.1007/s10664-015-9415-3
![]() |
[35] | Z. Wei, T. Shen, X. Chen, Just-in-time defect prediction technology based on interpretability technology, in 2021 8th International Conference on Dependable Systems and Their Applications (DSA), (2021), 78-89. https://doi.org/10.1109/DSA52907.2021.00017 |
[36] | L. Ma, F. Zhang, J. Sun, M. Xue, B. Li, F. J. Xu, et al., Deepmutation: Mutation testing of deep learning systems, in 2018 IEEE 29th International Symposium on Software Reliability Engineering (ISSRE), (2018), 100-111. https://doi.org/10.1109/ISSRE.2018.00021 |
[37] | G. Petrovic, M. Ivankovic, B. Kurtz, P. Ammann, R. Just, An industrial application of mutation testing: Lessons, challenges, and research directions, in 2018 IEEE International Conference on Software Testing, Verification and Validation Workshops (ICSTW), (2018), 47-53. https://doi.org/10.1109/ICSTW.2018.00027 |
[38] | Q. Hu, L. Ma, X. Xie, B. Yu, Y. Liu, J. Zhao, DeepMutation++: A mutation testing framework for deep learning systems, in 2019 34th IEEE/ACM International Conference on Automated Software Engineering (ASE), (2019), 1158-1611. https://doi.org/10.1109/ASE.2019.00126 |
[39] |
P. Gómez-Abajo, E. Guerra, J. D. Lara, M. G. Merayo, Wodel-Test: a model-based framework for language-independent mutation testing, Software Syst. Model., 20 (2021), 767-793. https://doi.org/10.1007/s10270-020-00827-0 doi: 10.1007/s10270-020-00827-0
![]() |
[40] | L. Chen, L. Zhang, Speeding up mutation testing via regression test selection: An extensive study, in 2018 IEEE 11th International Conference on Software Testing, Verification and Validation (ICST), (2018), 58-69. https://doi.org/10.1109/ICST.2018.00016 |
[41] | N. Humbatova, G. Jahangirova, P. Tonella, DeepCrime: mutation testing of deep learning systems based on real faults, in Proceedings of the 30th ACM SIGSOFT International, Symposium on Software Testing and Analysis, (2021), 67-78. https://doi.org/10.1145/3460319.3464825 |
[42] | K. Moran, M. Tufano, C. Bernal-Cárdenas, M. Linares-Vásquez, G. Bavota, C. Vendome, et al., MDroid+: A mutation testing framework for android, in 2018 IEEE/ACM 40th International Conference on Software Engineering: Companion (ICSE-Companion), (2018), 33-36. https://doi.org/10.1145/3183440.3183492 |
[43] | Z. Li, H. Wu, J. Xu, X. Wang, L. Zhang, Z. Chen, MuSC: A tool for mutation testing of ethereum smart contract, in 2019 34th IEEE/ACM International Conference on Automated Software Engineering (ASE), (2019), 1198-1201. https://doi.org/10.1109/ASE.2019.00136 |
[44] | D. Mao, L. Chen, L. Zhang, An extensive study on cross-project predictive mutation testing, in 2019 12th IEEE Conference on Software Testing, Validation and Verification (ICST), (2019), 160-171. https://doi.org/10.1109/ICST.2019.00025 |
[45] | D. Cheng, C. Cao, C. Xu, X. Ma, Manifesting bugs in machine learning code: An explorative study with mutation testing, in 2018 IEEE International Conference on Software Quality, Reliability and Security (QRS), (2018), 313-324. https://doi.org/10.1109/QRS.2018.00044 |
[46] | S. Lee, D. Binkley, R. Feldt, N. Gold, S. Yoo, Causal program dependence analysis, preprint, arXiv: 2104.09107. https://doi.org/10.48550/arXiv.2104.09107 |
[47] | S. Oh, S. Lee, S. Yoo, Effectively sampling higher order mutants using causal effect, in 2021 IEEE International Conference on Software Testing, Verification and Validation Workshops, (2021), 19-24. https://doi.org/10.1109/ICSTW52544.2021.00017 |
[48] |
X. Cai, Y. Niu, S. Geng, J. Zhang, Z. Cui, J. Li, et al., An under-sampled software defect prediction method based on hybrid multi-objective cuckoo search, Concurrency Comput.: Pract. Exper., 32 (2020), e5478. https://doi.org/10.1002/cpe.5478 doi: 10.1002/cpe.5478
![]() |
[49] | A. Rahman, J. Stallings, L. Williams, Defect prediction metrics for infrastructure as code scripts in DevOps, in Proceedings of the 40th International Conference on Software Engineering: Companion Proceeedings, (2018), 414-415. https://doi.org/10.1145/3183440.3195034 |
[50] | A. Amar, P. C. Rigby, Mining historical test logs to predict bugs and localize faults in the test logs, in 2019 IEEE/ACM 41st International Conference on Software Engineering. (2019), 140-151. https://doi.org/10.1109/ICSE.2019.00031 |
[51] |
P. Sangameshwar, B. Ravindran, Predicting software defect type using concept-based classification, Empirical Software Eng., 25 (2020), 1341-1378. https://doi.org/10.1142/S1469026820500054 doi: 10.1142/S1469026820500054
![]() |
[52] |
S. Wang, T. Liu, J. Nam, L. Tan, Deep semantic feature learning for software defect prediction, IEEE Trans. Software Eng., 46 (2018), 1267-1293. https://doi.org/10.1109/TSE.2018.2877612 doi: 10.1109/TSE.2018.2877612
![]() |
[53] |
X. Yin, L. Liu, H. Liu, Q. Wu, Heterogeneous cross-project defect prediction with multiple source projects based on transfer learning, Math. Biosci. Eng., 17 (2020), 1020-1040. https://doi.org/10.3934/mbe.2020054 doi: 10.3934/mbe.2020054
![]() |
[54] |
L. Qiao, X. Li, Q. Umer, P. Guo, Deep learning based software defect prediction, Neurocomputing, 385 (2020), 100-110. https://doi.org/10.1016/j.neucom.2019.11.067 doi: 10.1016/j.neucom.2019.11.067
![]() |
[55] |
F. Wu, X. Y. Jing, Y. Sun, J. Sun, L. Huang, F. Cui, et al., Cross-project and within-project semisupervised software defect prediction: A unified approach, IEEE Trans. Reliab., 67 (2018), 581-597. https://doi.org/10.1109/TR.2018.2804922 doi: 10.1109/TR.2018.2804922
![]() |
[56] |
D. L. Miholca, G. Czibula, I. G. Czibula, A novel approach for software defect prediction through hybridizing gradual relational association rules with artificial neural networks, Inf. Sci., 441 (2018), 152-170. https://doi.org/10.1016/j.ins.2018.02.027 doi: 10.1016/j.ins.2018.02.027
![]() |
[57] |
A. Majd, M. Vahidi-Asl, A. Khalilian, P. Poorsarvi-Tehrani, H. Haghighi, SLDeep: Statement-level software defect prediction using deep-learning model on static code features, Expert Syst. Appl., 147 (2020), 113156. https://doi.org/10.1016/j.eswa.2019.113156 doi: 10.1016/j.eswa.2019.113156
![]() |
[58] | G. G. Cabral, L. L. Minku, E. Shihab, S. Mujahid, Class imbalance evolution and verification latency in just-in-time software defect prediction, in 2019 IEEE/ACM 41st International Conference on Software Engineering (ICSE), (2019), 666-676. https://doi.org/10.1109/ICSE.2019.00076 |
[59] | A. Perez, R. Abreu, A. Deursen, A test-suite diagnosability metric for spectrum-based fault localization approaches, in 2017 IEEE/ACM 39th International Conference on Software Engineering, (2017), 654-664. https://doi.org/10.1109/ICSE.2017.66 |
[60] | X. Li, W. Li, Y. Zhang, L. Zhang, Deepfl: Integrating multiple fault diagnosis dimensions for deep fault localization, in Proceedings of the 28th ACM SIGSOFT International Symposium on Software Testing and Analysis, (2019), 169-180. https://doi.org/10.1145/3293882.3330574 |
[61] |
A. Zakari, S. P. Lee, R. Abreu, B. H. Ahmed, R. A. Rasheed, Multiple fault localization of software programs: A systematic literature review, Inf. Software Technol., 124 (2020), 106312. https://doi.org/10.1016/j.infsof.2020.106312 doi: 10.1016/j.infsof.2020.106312
![]() |
[62] | Z. Li, Y. Wu, Y. Liu, An empirical study of bug isolation on the effectiveness of multiple fault localization, in 2019 IEEE 19th International Conference on Software Quality, Reliability and Security (QRS), (2019), 18-25. https://doi.org/10.1109/QRS.2019.00016 |
[63] | H. L. Ribeiro, R. P. A. de Araujo, M. L. Chaim, H. A. de Souza, F. Kon, Jaguar: A spectrum-based fault localization tool for real-world software, in 2018 IEEE 11th International Conference on Software Testing, Verification and Validation (ICST), (2018), 404-409. https://doi.org/10.1109/ICST.2018.00048 |
[64] | K. Liu, A. Koyuncu, T. F. Bissyandé, D. Kim, J. Klein, Y. Le Traon, You cannot fix what you cannot find! An investigation of fault localization bias in benchmarking automated program repair systems, in 2019 12th IEEE conference on software testing, validation and verification (ICST), (2019), 102-113. https://doi.org/10.1109/ICST.2019.00020 |
[65] |
D. Zou, J. Liang, Y. Xiong, M. D. Ernst, L. Zhang, An empirical study of fault localization families and their combinations, IEEE Trans. Software Eng., 47 (2019), 332-347. https://doi.org/10.1109/TSE.2019.2892102 doi: 10.1109/TSE.2019.2892102
![]() |
[66] |
J. Kim, J. Kim, E. Lee, Variable-based fault localization, Inf. Software Technol., 107 (2019), 179-191. https://doi.org/10.1016/j.infsof.2018.11.009 doi: 10.1016/j.infsof.2018.11.009
![]() |
[67] |
Y. Kim, S. Mun, S. Yoo, M. Kim, Precise learn-to-rank fault localization using dynamic and static features of target programs, ACM Trans. Software Eng. Method. (TOSEM), 28 (2019), 1-34. https://doi.org/10.1145/3345628 doi: 10.1145/3345628
![]() |
[68] | Defects4J, 2022. available from: https://github.com/rjust/defects4j. |
1. | Abdul Alamin, Sankar Prasad Mondal, Kunal Biswas, Shariful Alam, 2020, chapter 6, 9781799837411, 120, 10.4018/978-1-7998-3741-1.ch006 | |
2. | Zhimin Han, Yi Wang, Jinde Cao, Impact of contact heterogeneity on initial growth behavior of an epidemic: Complex network-based approach, 2023, 451, 00963003, 128021, 10.1016/j.amc.2023.128021 |