Telematika – Vol. 14, No. 1, February (2021) pp.

13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

Available Online at:

http://ejournal.amikompurwokerto.ac.id/index.php/telematika/

Telematika
Accreditated SINTA “2” Kemenristek/BRIN, No. 85/M/KPT/2020

Local Sensitivity Analysis of COVID-19 Epidemic

with Quarantine and Isolation using Normalized Index

Muhammad Abdurrahman Rois1, Trisilowati2, Ummu Habibah3

1,2,3
Department of Mathematics, Faculty of Mathematics and Natural Sciences, Brawijaya University
E-mail: roizmuhammad.math@gmail.com1, trisilowati@ub.ac.id2, ummu_habibah@ub.ac.id3

ARTICLE INFO ABSTRACT

History of the article: This study discusses the sensitivity analysis of parameters, namely the COVID-
Received December 19, 2020 19 model, by dividing the population into seven subpopulations: susceptible,
Revised January 12, 2021
Accepted February 25, 2021
exposed, symptomatic infection, asymptomatic infection, quarantine, isolation,
and recovered. The solution to the ordinary differential equation for the COVID-
19 model using the fourth-order Runge-Kutta numerical method explains that
Keywords:
COVID-19 model, COVID-19 is endemic, as evidenced by the basic reproduction number (𝑅0 ) of
Runge-Kutta fourth-order, 7.5. It means 1 individual can infect 7 to 8 individuals. Then 𝑅0 is calculated using
Next-generation matrix, the next-generation matrix method. Based on the value of 𝑅0 , a parameter
Sensitivity analysis sensitivity analysis is implemented to specify the most influential parameters in
the spread of the COVID-19 outbreak. This can provide input on the selection of
Correspondece:
Telephone: +62 81914430369 appropriate control measures to solve the epidemic from COVID-19. The results
E-mail: of the sensitivity analysis are the parameters that have the most influence on the
roizmuhammad.math@gmail.com model, namely Λ, 𝑞1 , 𝑞2 , 𝜎, 𝛽2 , 𝛾3 , 𝛽3 , 𝛽1 𝑑, 𝛾1 , 𝜔, 𝜃, 𝜇.

INTRODUCTION
COVID-19 first appeared and was identified in Wuhan, China, around December 2019 (Chen et al.,
2020), then spread to various countries rapidly through people who had a history of travel to Wuhan (Tang
et al., 2020; WHO, 2020; Yousefpour, Jahanshahi, & Bekiros, 2020). On October 2, 2020, COVID-19
caused a total of 34,819,505 cases in the world to be positively infected. The latest information was on
October 2, 2020, more than 34 million infected cases in the world, with 1,032,741 deaths have been
reported. In Indonesia's first case on March 2, 2020, there were 2 people who were infected, namely a
mother and child who were suspected of contracting from a Japanese citizen (Sumartiningtyas, 2020).
While on October 2, 2020, data from the web Worldometer that Indonesia ranks 23 out of 215 countries
reported as infected with 295,499 confirmed cases, 10,972 deaths, and 221,340 recoveries. The number of
COVID-19 cases in Indonesia continues to increase every day. The government has made various efforts
to control COVID-19, such as social distancing, tracing, and working from home. Apart from some areas,
the government has also imposed very stick territorial restrictions.
The mathematical model that can describe the spread of disease is SIR (Susceptible-Infected-
Recovered) model. SIR model was first introduced by Kermack and McKendrick (1927), then became a
source of reference and played a role in the development of mathematics about the transmission of
infectious diseases (Müller & Kuttler, 2015; Murray, 2002). Modeling of the COVID-19 case has been
widely carried out with various models starting from research conducted by Soewono (2020) using SEIR
model, SEIAR model (Belgaid, Helal, & Venturino, 2020), SEIHR model (Zeb, Alzahrani, Erturk, &
http://dx.doi.org/10.35671/telematika.v14i1.1191 13
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

Zaman, 2020), and SEIAQHR model (Jia et al., 2020). This study uses the SEIAQHR model based on
conditions in Indonesia and a combination of research from Belgaid et al. (2020), Jia et al. (2020), and Zeb
et al. (2020). The steps taken in this research are first, the construction of the COVID-19 model with
quarantine and isolation is carried out. Second, 𝑅0 value is obtained which is calculated by the next-
generation matrix method. Furthermore, a sensitivity analysis is implemented to show the most significant
parameters in the spread of COVID-19.

RESEARCH METHODS
1. Tools and Materials
In this study, the tools and materials used are as follows: (i) MATLAB 2017. (ii) Computer with
processor specifications intel i5-8250U, RAM 8GB, and Windows 10 64-bit operation.
2. Research Flow
In this study, the following steps were carried out:

Figure 1. Flowchart of the COVID-19 model sensitivity analysis research

The explanation of the research steps are as follows:

a. Construction of COVID-19 model
This model uses the SEIAQHR model based on conditions in Indonesia and a combination of
research from Belgaid et al., Jia et al., and Zeb et al. in 2020 about the COVID-19 model.
b. Basic reproduction numbers (𝑅0 )
𝑅0 is the average number of newly infected individuals caused by one infected individual in a
susceptible population. 𝑅0 is used to determine the spread of disease and predict a population can
endanger or not. The conditions that arise are among the following possibilities (Heffernan,
Smith, & Wahl, 2005): (i) If 𝑅0 < 1, it means that an infected individual can transmit the disease
on average less than one newly infected individual, so it can be predicted that the infection will
disappear and there will be no spread of disease, or it is called disease-free. (ii) If 𝑅0 > 1, it means

http://dx.doi.org/10.35671/telematika.v14i1.1191 14
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

that an infected individual can transmit the disease on average more than one newly infected
individual, causing it to easily spread the disease.
Furthermore, the next-generation matrix (NGM) is one method to determining R0 . In the

NGM method, the compartment model used is the infected compartment (Brauer & Castillo-
Chavez, 2012). Furthermore, the infected compartment model can be written as:
dx
 f  v. (1)
dt
So, obtained
f  E0  v  E0 
F ,V  . (2)
x x

NGM is defined as M  FV 1 and 𝑅0 can be obtained from R0    M  , where   M 

is the spectral radius of the matrix M , which is the largest modulus of the eigenvalues of the
matrix M .
c. Parameter sensitivity analysis
Parameter sensitivity analysis is used to identify the effect of each parameter on disease
spread based on 𝑅0 , this can provide input in the selection of appropriate control measures to
prevent an outbreak from COVID-19. The sensitivity analysis is divided into two based on Ingalls
(2012) is: (i) Global sensitivity analysis, which discusses the wide variation in a parameter value
which is often implemented using the Monte-Carlo method, and (ii) Local sensitivity analysis,
namely discussing local variations around, using the one-at-a-time (OAT) technique, which is the
most basic method with partial differentiation, in which various value parameters are taken one
by one.
In the local sensitivity analysis using the normalized sensitivity index. The normalized sensitivity
index is defined as follows (Chitnis, Hyman, & Cushing, 2008):
V p
CVp  , (3)
p V
Where V is the variable to be analyzed, and p is the parameter.
d. The fourth-order Runge-Kutta
The fourth-order Runge-Kutta is a numerical method commonly used to solve ordinary
𝑑𝑦
differential equations (Lenhart and Workman, 2007). In the initial value problem, if = 𝑓(𝑡, 𝑦(𝑡))
𝑑𝑡

and 𝑦(𝑡) is known. The value of 𝑦(𝑡 + ℎ) can be approximated by

h
y t  h   y t    n1  2n2  2n3  n4  , (4)
6
With:
n1  f  t , y  t   ,
 1 1 
n2  f  t  h, y  t   hn1  ,
 2 2 
(5)
 1 1 
n3  f  t  h, y  t   hn2  ,
 2 2 
n4  f  t  h, y  t   hn3  ,

Where:

http://dx.doi.org/10.35671/telematika.v14i1.1191 15
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

𝑛1 is use 𝑥 for the slope at the beginning of the interval,

𝑛2 is use 𝑥 and 𝑚1 for the slope at the midpoint of the interval,
𝑛3 is again the slope at the midpoint but now using 𝑥 and 𝑚2 ,
𝑛4 is use 𝑥 and 𝑚3 for the slope at the end of the interval.

RESULTS AND DISCUSSION

The construction of the COVID-19 model with quarantine and isolation is given as follows:
dS
   S  1 E   2 I   3 A    q1  ,
dt
dE
 S  1 E   2 I   3 A   E      q2  ,
dt
dI
  E   A  I   1    d  ,
dt
dA
 1     E  A   3      , (6)
dt
dQ
 q1 S  q2 E  Q     2  r2  ,
dt
dH
  1 I   2 Q   3 A  H    r1  ,
dt
dR
 r1 H  r2 Q   R.
dt
Description
 : Human recruitment rate,
1 : Contact rate of susceptible subpopulations with exposed,

 2 : Subpopulation's contact rate is susceptible to symptomatic,

 3 : Subpopulation's contact rate is susceptible to asymptomatic,

 : Rate of natural death,
d : Rate of death due to COVID-19,
q1 : Rate of susceptible individuals quarantined,

q 2 : Rate of exposed individuals quarantined,

 : Rate of return of individuals to susceptible subpopulations after quarantine,

 1 : Isolation rate from symptomatic subpopulations,
 2 : Isolation rate from quarantine subpopulation,

 3 : Isolation rate from asymptomatic subpopulation,

 : Rate of asymptomatic individuals becomes symptomatic,

r1 : Recovery rate after isolation,

r2 : Recovery rate after quarantine,

 : Rate of progression from exposed to symptomatic,

 : Proportion of becoming infected is symptomatic.
Based on the system of equation (6), 𝑅0 is obtained using the next-generation matrix as follows:

http://dx.doi.org/10.35671/telematika.v14i1.1191 16
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

   2  h4       3 1    
R0    M    1    (7)
h1h2  h3 h4 h4 
with
ℎ1 =   q1 , ℎ2 =     q2 , ℎ3 =  1    d , dan ℎ4 =  3    .

1. Preliminary data
The parameters used in research in the Indonesia region are as follows:
Table 1. Parameters used in the COVID-19 model
Parameter
Parameter Description Source
Value
𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙
Λ 1.685 Human recruitment rate ( ) Assumed
𝑑𝑎𝑦×1,00
1 (Aldila, Ndii, &
𝜇 3.9139 Χ 10−5 Natural date rate ( )
𝑑𝑎𝑦 Samiadji, 2020)
Rate of progression from exposed to
𝜎 0.196 Assumed
symptomatic
Rate of asymptomatic individuals becomes (Aldila, Ndii, &
𝜃 0.01 1
symptomatic ( ) Samiadji, 2020)
𝑑𝑎𝑦
Proportion of becoming infected is (Aldila, Ndii, &
𝜔 0.4
symptomatic Samiadji, 2020)
(Sasmita, Ikhwan,
Suyanto, &
𝑑 0.087 Death rate due to COVID-19
Chongsuvivatwong,
2020)
Rate of susceptible individuals quarantined
𝑞1 0.09 1 Assumed
( )
𝑑𝑎𝑦
Rate of exposed individuals quarantined
𝑞2 0.1 1 Assumed
( )
𝑑𝑎𝑦
Contact rate of susceptible subpopulations
𝛽1 0.01 1 Assumed
with exposed ( )
𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙×𝑑𝑎𝑦
Subpopulation's contact rate is susceptible to
𝛽2 0.4 1 Assumed
symptomatic ( )
𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙×𝑑𝑎𝑦
Subpopulation's contact rate is susceptible to
𝛽3 0.4 1 Assumed
asymptomatic ( )
𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙×𝑑𝑎𝑦
Table 2. Next parameters used in the COVID-19 model
Parameter
Parameter Description Source
Value
Isolation rate from symptomatic (Aldila, Ndii, &
𝛾1 0.083 1
subpopulations ( ) Samiadji, 2020)
𝑑𝑎𝑦
Isolation rate from quarantine subpopulation (Aldila, Ndii, &
𝛾2 0.01 1
( ) Samiadji, 2020)
𝑑𝑎𝑦
Isolation rate from the asymptomatic (Aldila, Ndii, &
𝛾3 0.2435 1
subpopulation ( ) Samiadji, 2020)
𝑑𝑎𝑦
1 (Aldila, Ndii, &
𝑟1 0.1 Recovery rate after isolation ( )
𝑑𝑎𝑦 Samiadji, 2020)
1 (Aldila, Ndii, &
𝑟2 0.125 Recovery rate after quarantine ( )
𝑑𝑎𝑦 Samiadji, 2020)

2. Model completion
Using Runge Kutta Formulas in equation (4) to find the next value from equations (6) as follows:

http://dx.doi.org/10.35671/telematika.v14i1.1191 17
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

1. %apply runge kutta formulas of equation (4)

2. for i=1:m1-1
3. n1=h1*fungsi(t1(i),y(i,:)); %Write equation (5)
4. n2=h1*fungsi(t1(i)+1/2*h1,y(i,:)+0.5*n1); %Write equation (5)
5. n3=h1*fungsi(t1(i)+1/2*h1,y(i,:)+0.5*n2); %Write equation (5)
6. n4=h1*fungsi(t1(i)+h1,y(i,:)+n3); %Write equation (5)
7. y(i+1,:)=y(i,:)+1/6*(n1+2*n2+2*n3+n4); %Find value
8. End

and the function program of a system (6) is presented as follows:

1. function dz=fungsi(t,z)
2. %call value
3. S=y(1);
4. E=y(2);
5. I=y(3);
6. A=y(4);
7. Q=y(5);
8. H=y(6);
9. R=y(7);
10. %--------------------------------------------
11. %Write equation (6)
12. dS=Lambda-S*(beta1*E+beta2*I+beta3*A+miu+q1);
13. dE=S*(beta1*E+beta2*I+beta3*A)-E*(miu+sigma+q2);
14. dI=sigma*w*E+theta*A-I*(gamma1+miu+d);
15. dA=(1-w)*sigma*E-A*(gamma3+theta+miu);
16. dQ=q1*S+q2*E-Q*(miu+gamma2+r2);
17. dH=gamma1*I+gamma2*Q+gamma3*A-H*(miu+r1);
18. dR=r1*H+r2*Q-miu*R;
19. dy=[dS dE dI dA dQ dH dR];

Furthermore, Figure 2 (a) - (g) is obtained by the following command:

1. plot(tW,W(:,1),'-green','LineWidth',1) %plot figure (a)
2. plot(tW,W(:,2),'-red','LineWidth',1) %plot figure (b)
3. plot(tW,W(:,3),'-y','LineWidth',1) %plot figure (c)
4. plot(tW,W(:,4),'-m','LineWidth',1) %plot figure (d)
5. plot(tW,W(:,5),'-b','LineWidth',1) %plot figure (e)
6. plot(tW,W(:,6),'-black','LineWidth',1) %plot figure (f)
7. plot(tW,W(:,7),'-c','LineWidth',1) %plot figure (g)

Based on parameters in Table 1 and Table 2, it can be seen that the susceptible (S) is significantly
reduced in Figure 2 (a). Then Figures 2 (b), (c), (d) show exposed (E), symptomatic (I), and asymptomatic
(A) up significantly and down significantly, but in a certain period until the end of the time, there are still
individuals exposed, symptomatic and asymptomatic. Therefore, endemic diseases still epidemic during
this period. Figures 2 (e), (f) show the quarantine (Q) and isolation (H) subpopulations showing significant
increases and decreases due to overcoming the spike in the exposed, symptomatic and asymptomatic
subpopulations. Furthermore, the subpopulation of quarantine and isolation after a certain time until the
end of time remains. The recovered subpopulation (R) in Figure 2 (g) shows that the recovery overtime is
increasing.

http://dx.doi.org/10.35671/telematika.v14i1.1191 18
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

Figure 2. Numerical solutions for the COVID-19 model with quarantine and isolation

3. Sensitivity analysis
The sensitivity analysis is measured using a sensitivity index. The sensitivity index makes it possible
to measure changes in the variable 𝑅0 when one parameter value is increased or decreased while the other
parameter values are maintained. A positive sensitivity index indicates that if the value of a parameter is
decreased or increased by 10%, then the value of 𝑅0 will decrease or increase by 10%. Conversely, a

http://dx.doi.org/10.35671/telematika.v14i1.1191 19
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

negative sensitivity index indicates that if the value of a parameter is decreased or increased, the value of
𝑅0 will increase or decrease. The sensitivity analysis is obtained by the following command
1. %write R0 of equation (7)
2. R0=a*(beta1+beta2*sigma*((gama3+t+miu)*w-
t*w+t)/((gama1+miu+d)*(gama3+t+miu))+beta3*sigma*(1-
w)/(gama3+t+miu))/((miu+q1)*(miu+sigma+q2));
3. %------------------------------------------------------------------%
4. %based on equation (3), partial derivative of R0 for each parameter
5. beta1b=diff(R0,beta1).*(beta1/R0); %derived to beta1
6. beta2b=diff(R0,beta2).*(beta2/R0); % derived to beta2
7. beta3b=diff(R0,beta3).*(beta3/R0); % derived to beta3
8. sigmab=diff(R0,sigma).*(sigma/R0); % derived to sigma
9. wb=diff(R0,w).*(w/R0); % derived to omega/w
10. tb=diff(R0,t).*(t/R0); derived to theta
11. miub=diff(R0,miu).*(miu/R0); % derived to miu
12. q1b=diff(R0,q1).*(q1/R0); % derived to q1
13. q2b=diff(R0,q2).*(q2/R0); % derived to q2
14. gama1b=diff(R0,gama1).*(gama1/R0); % derived to gamma1
15. gama3b=diff(R0,gama3).*(gama3/R0); % derived to gamma3
16. db=diff(R0,d).*(d/R0); % derived to d

Figure 3. Normalized sensitivity index

The normalized sensitivity index of 𝑅0 is related to the parameters shown in Figure 3. Parameters that
have a positive index, namely Λ, 𝛽1 , 𝛽2 , 𝛽3 , 𝜎, 𝜔, and 𝜃 are a positive effect on 𝑅0 . It means an increase in
value Λ, 𝛽1 , 𝛽2 , 𝛽3 , 𝜎, 𝜔, 𝜃 can increase the 𝑅0 value or cause an outbreak. Furthermore, the increase of
parameters whose sensitivity index is negative is 𝑞1 , 𝑞2 , 𝛾3 , 𝛾1 , 𝑑 and 𝜇 a negative effect of minimizing the
spread of disease. Furthermore, to specify the effect of changing parameter values on 𝑅0 , the value of each
parameter is increased or decreased. Suppose that each parameter is increased or decreased by 10%. Then
the effect of changes in parameter values that are increased or decreased by 10% on 𝑅0 is shown in Table
3, sorted from most sensitive to less sensitive as follows.
Table 3. Effect of changes in parameter values
𝑅0 = 7.5
No Parameter
p + 10% p − 10%
1 Λ 8.2741 6.7697
2 q1 6.8383 8.3572
3 q2 6.9417 8.2079
4 σ 7.9868 7.0415

http://dx.doi.org/10.35671/telematika.v14i1.1191 20
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

5 β2 7.8274 7.2163
6 γ3 7.2526 7.8483
7 β3 7.8121 7.2317
8 β1 7.6783 7.3654
9 d 7.3731 7.6867
10 γ1 7.3797 7.6787
11 ω 7.6055 7.4382
12 θ 7.5268 7.5169
13 μ 7.5212 7.5226

After doing the analysis, the parameters are increased or decreased by 10%. Then an example is given
a simulation of the effect of increasing or decreasing parameters applied to all subpopulations, as shown in
Figure 4 and Figure 5 below.

Figure 4. Simulation of the effect of changes in parameter values

http://dx.doi.org/10.35671/telematika.v14i1.1191 21
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

Figure 5. Next simulation of the effect of changes in parameter values

Figure 4 and Figure 5 show that the increase and decrease in parameters that affect the spread of
COVID-19 will affect each subpopulation, especially in the exposed, symptomatic, and asymptomatic
subpopulations. This also affects the quarantine and isolation subpopulations in response to an increase or
decrease in infections. The reason is, these parameters have proven to be the most influential in increasing
the outbreak and reducing outbreak COVID-19.
In this study, from Table 3, one parameter is taken to be used as an example of a graph of these
parameters' influence. The parameter is taken 𝑞1 , and the parameter effect graph is shown in Figure 4 and
Figure 5. Increasing and decreasing the 𝑞1 parameter by 10% can change the value of 𝑅0 . Because 𝑞1 has
a negative effect, it means that if the parameter value of 𝑞1 is increased by only 10%, the value of 𝑅0 will
decrease, but if the value of the parameter 𝑞1 is reduced by 10%, the value of 𝑅0 will increase.
Furthermore, the parameters obtained using the normalized sensitivity index in Figure 3 have
positive and negative values. The parameters that have a positive effect are Λ, 𝛽1 , 𝛽2 , 𝛽3 , 𝜎, 𝜔, 𝜃 and it means
that if the parameters that have a positive effect are increased or decreased by only 10%, the 𝑅0 the value
will increase or decrease. Then, the parameters that have a negative effect are 𝑞1 , 𝑞2 , 𝛾3 , 𝛾1 , 𝑑, 𝜇 and it
means that if the parameters have a negative effect are increased or decreased by 10%, the value of 𝑅0 will
decrease or increase. Therefore, to control the spread of COVID-19 decreases, we can increase negative
parameters and decrease positive parameters in Figure 3.

http://dx.doi.org/10.35671/telematika.v14i1.1191 22
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

CONCLUSIONS AND RECOMMENDATIONS

The results of completing the COVID-19 model with quarantine and isolation are in equation (4),
and with the initial values in Table 1 using the fourth-order Runge-Kutta method, it is found that the
COVID-19 disease is endemic (Figure 2 and Figure 3) and is proven by the value 𝑅0 = 7.5. After
conducting a sensitivity analysis, several parameters that have the most influence in the spread of COVID-
19 are obtained, namely 𝛬, 𝑞1 , 𝑞2 , 𝜎, 𝛽2 , 𝛾3 , 𝛽3 , 𝛽1 , 𝑑, 𝛾1 , 𝜔, 𝜃, 𝜇. From the sensitivity analysis, it can provide
to selection of effective control measures in equation (4) to reduce the severity of the outbreak or eliminate
the COVID-19 outbreak.
This study uses parameters from previous studies, and some are assumed so that if applied to this
model it is not necessarily 100% valid, it is necessary to estimate parameters that are considered not
knowing the exact value such as 𝛬, 𝛽1 , 𝛽2 , 𝛽3 , 𝜎, 𝑞1 , 𝑞2 , 𝛾1 , 𝛾2 , 𝛾3 . Then several algorithms can also be
compared to complete parameter estimates, for example, the trust-region algorithm (subroutine
fminsearch), genetic algorithms, PSO (Particle Swarm Optimization) algorithms, and others. Furthermore,
it can also be developed to take control measures to solve the COVID-19 disease outbreak based on
parameter sensitivity analysis.

REFERENCES
-. (2020). Report Coronavirus Cases. https://www.worldometers.info/coronavirus/

Aldila, D., Ndii, M. Z., & Samiadji, B. M. (2020). Optimal Control on COVID-19 Eradication Program in
Indonesia under the Effect of Community Awareness. Mathematical Biosciences and Engineering,
17(6), 6355–6389. doi: 10.3934/mbe.2020335

Belgaid, Y., Helal, M., & Venturino, E. (2020). Analysis of a Model for Coronavirus Spread. Mathematics,
8(5), 1–30. https://doi.org/10.3390/math8050820

Brauer, F., & Castillo-Chavez, C. (2012). Mathematical Models in Population Biology and Epidemiology.
The American Mathematical Monthly (Second Edi., Vol. 110). New York: Springer-Verlag New
York. DOI: 10.1007/978-1-4614-1686-9\_7

Chen, T. M., Rui, J., Wang, Q. P., Zhao, Z. Y., Cui, J. A., & Yin, L. (2020). A Mathematical Model for
Simulating the Phase-Based Transmissibility of a Novel Coronavirus. Infectious Diseases of
Poverty, 9(1), 1–8. Infectious Diseases of Poverty. https://doi.org/10.1186/s40249-020-00640-3

Chitnis, N., Hyman, J. M., & Cushing, J. M. (2008). Determining Important Parameters in the Spread of
Malaria through the Sensitivity Analysis of a Mathematical Model. Bulletin of Mathematical
Biology, 70(5), 1272. DOI: 10.1007/s11538-008-9299-0

Heffernan, J. M., Smith, R. J., & Wahl, L. M. (2005). Perspectives on the Basic Reproductive Ratio. Journal
of the Royal Society Interface, 2(4), 281–293. doi: 10.1098/rsif.2005.0042

Ingalls, B. (2012). Mathematical Modelling in Systems Biology: An Introduction. Waterloo: University of

Waterloo.

Jia, J., Ding, J., Liu, S., Liao, G., Li, J., Duan, B., Wang, G., et al. (2020). Modeling the Control of COVID-
19: Impact of Policy Interventions and Meteorological Factors. Electronic Journal of Differential
Equations, 2020(23), 1–24.

Lenhart, S., & Workman, J. T. (2007). Optimal Control Applied to Biological Models. New York: Chapman
and Hall/CRC.

Müller, J., & Kuttler, C. (2015). Methods and Models in Mathematical Biology. Methods and Models in
Mathematical Programming. London: Springer-Verlag Berlin Heidelberg. DOI: 10.1007/978-3-
642-27251-6

http://dx.doi.org/10.35671/telematika.v14i1.1191 23
Telematika – Vol. 14, No. 1, February (2021) pp. 13-24 e-ISSN 2442-4528 | p-ISSN 1979-925X

Murray, J. D. (2002). Mathematical Biology I: An Introduction (3rd ed., Vol. 17). New York: Springer-
Verlag Berlin Heidelberg.

Sasmita, N. R., Ikhwan, M., Suyanto, S., & Chongsuvivatwong, V. (2020). Optimal Control on a
Mathematical Model to Pattern the Progression of Coronavirus Disease 2019 (COVID-19) in
Indonesia. Global Health Research and Policy, 5.

Soewono, E. (2020). On the Analysis of Covid-19 Transmission in Wuhan, Diamond Princess, and Jakarta-
cluster. Communication in Biomathematical Sciences, 3(1), 9–18.

Sumartiningtyas, H. K. N. (2020, March 2). Update Virus Corona 2 Maret: 89.212 Orang di 68 Negara
Terinfeksi. KOMPAS. Retrieved from
https://sains.kompas.com/read/2020/03/02/170200123/update-virus-corona-2-maret-89212-orang-
di-68-negara-terinfeksi?page=all

Tang, B., Wang, X., Li, Q., Bragazzi, N. L., Tang, S., Xiao, Y., & Wu, J. (2020). Estimation of the
Transmission Risk of the 2019-nCoV and Its Implication for Public Health Interventions. Journal
of Clinical Medicine, 9(2), 462. doi: 10.3390/jcm9020462

WHO. (2020). Pertimbangan-pertimbangan untuk karantina individu dalam konteks penanggulangan

penyakit coronavirus (COVID-19). Retrieved from https://www.who.int/docs/default-
source/searo/indonesia/covid19/who-2019-covid19-ihr-quarantine-2020-
indonesian.pdf?sfvrsn=31d7cbd8_2

Yousefpour, A., Jahanshahi, H., & Bekiros, S. (2020). Optimal Policies for Control of the Novel
Coronavirus Disease (COVID-19) Outbreak. Chaos, Solitons and Fractals, 136.
https://doi.org/10.1016/j.chaos.2020.109883

Zeb, A., Alzahrani, E., Erturk, V. S., & Zaman, G. (2020). Mathematical Model for Coronavirus Disease
2019 (COVID-19) Containing Isolation Class. BioMed Research International, 2020.
https://doi.org/10.1155/2020/3452402

http://dx.doi.org/10.35671/telematika.v14i1.1191 24