This research explains the applications of the algorithm that was designed to provide statistical support for

medical doctors. The support that was achieved from this research looks for an urgent interpretation of

parameters such as the rate of infection by COVID-19 and the rate of mortality because of the virus.

Furthermore, this research achieves prediction rates that were provided by a mathematical model that

observes and adapts real statistical data from other countries, where governments are trying to find

solutions for COVID-19 propagation. In order to get accuracy in prediction results, it was necessary to

analyse the statistical behaviour from China and other countries that returned to normal activities before

virus required confinement of the population inside homes. The viruses problematic growth and some

suggestions, on how to avoid deep complications in health and economy of people (for instance,

quarantine as the main response to attenuate advance of this virus) are offered.

The Biologist

(Lima)

The Biologist (Lima), 2021, vol. 19 (1), 19-28.

ORIGINAL ARTICLE / ARTÍCULO ORIGINAL

1Applied Biophysics, Institute for Physics, Technical University of Ilmeanu , Ilmenau 98693, Germany

2Pontiﬁcia Universidad Católica del Perú, Mechatronic Master Program and Energy Laboratory, Lima 32, Peru

3Department of Protocol and Investigation National Hospital Guillermo Almenara, Perú.

4 Northern (Artic) Federal University named after MV Lomonosov, Arkhangelsk Russia.

*Corresponding author: alan.calderon.@pucp.edu.pe

1,2 2 2 2,3

J. Alan Calderón Ch ; Julio Tafur Sotelo ; Benjamin Barriga Gamarra ; Julio Guevara Guevara ;

2,4 2 2

John Lozano Jauregui ; Juan Lengua Arteaga & Gonzalo Solano

ABSTRACT

Keywords: COVID-19 propagation – statistical data – Mathematical modeling

The Biologist (Lima)

ISSN Versión Impresa 1816-0719

ISSN Versión en linea 1994-9073 ISSN Versión CD ROM 1994-9081

doi:10.24039/rtb2021191878

PREDICTIVE ALGORITHM ANALYSIS FOR OPTIMAL PREVENTIONS IN TIME OF

CORONAVIRUS (COVID-19) DURING QUARANTINE

ANÁLISIS PREDICTIVO DEL ALGORITMO PARA UNA PREVENCIÓN ÓPTIMA EN EL

TIEMPO DE CORONAVIRUS (COVID-19) DURANTE LA CUARENTENA

https://orcid.org/0000-0002-6486-5105

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

RESUMEN

Palabras clave: Propagación de COVID–19 – datos estadísticos – modelamiento matemático

En esta investigación se explica las aplicaciones del algoritmo que fue propuesto para proporcionar apoyo

estadístico para los médicos. El apoyo que se realizó en esta investigación busca una urgente

interpretación de parámetros como la tasa de personas infectadas por COVID-19 y la tasa de personas

fallecidas a causa de este virus. Además, esta investigación logra predecir las tasas que fueron

proporcionadas por un modelo matemático que observa y adapta datos estadísticos reales de otros países

donde están tratando de encontrar soluciones contra la propagación del COVID-19. Esto implica que, con

el fin de obtener precisión en los resultados de la predicción, fue necesario analizar cuál fue el

comportamiento estadístico de China y otros países que volvieron a la normalidad de sus actividades, tal

como era antes de que el virus impusiera a la población a permanecer en sus hogares. Por otro lado, se

resume el crecimiento problemático del virus y algunas sugerencias de cómo evitar complicaciones

profundas en la salud y la economía de las personas (por ejemplo, los días de cuarentena, como principal

respuesta para atenuar el avance de este virus).

INTRODUCTION

In March 6, 2020, the first infected person

(Coronavirus en el Perú, 2020), was detected in

Peru (Figure 1, letter “A”). Therefore, there was

many recommendations given by Peruvian

government such as cleaning and keeping rest in

home, as it was made by the international

community, which is crossing the same task.

Nevertheless, it was expected exponential growth

of the infected quantity of people curve that was the

reason why the government decided to declare days

of quarantine from March 16, 2020 (El Peruano,

2020).

For this reason, data (as all statistical data about

COVID-19 in Peru obtained from official and

public information of Peruvian Health Ministry,

MINSA) was processed to show the behaviour of

infected quantity growth as the dependence of

current days and quarantine days.

Figure 1. The curve of the quantity of infected people as the dependence on 18 current days and 3 quarantine days.

Calderón-Chavarri et al.

Hence, the blue colour curve of Figure 1 shows the

increase of people, who were infected by the virus

even though after 10 days of finding the first

infected person in Peru, the “Quarantine days”

started (as it is possible to see from the letter “B” in

the same Figure).In order to give a mathematical

formalism for the statistical data that was achieved,

this research analyse a general polynomial (it is

showed in equation 1), which depends on its

coefficients (a, b) and derivatives P =d/dt (with

order n) as the consequence of the response Y, the

excitation variables u, internal variables x and error

e (all of them depend on time t) (Pearson, 1979;

Chowell et al., 2016; Chen et al., 2020).

This general polynomial model is known as

Modulating Functions technique that is very

adapted for many changes and its solution is

another polynomial model that was adjusted to the

desired response. Hence, it is possible to achieve

predictions that depend on adaptive coefficients, it

means a better accuracy to get estimations in

multivariable and non-linear models.

where solution error analysis e(t) is the error

described by the following equation, in which, V

keeps the Fourier series coefficients, indexes m, n,

k, a, moreover the parameters matrix θ.

In the following equation, α is the frequency

parameter function, C is the combination for

coefficients, a keeps the physical parameters and

n−j

(ikω) is a complex number for the corresponding

index:

MATERIALS AND METHODS For which, the nonlinear model for error analysis is

given by the following equation, in which, g is the

given function for parameters θ, E and F are

specified functions for the input variables u and

responses y, P are the fixed polynomials as

dependence on derivatives p = d/dt:

Therefore, the costing function that depends on the

function r of the complex conjugations is given by

the equation 5:

also, in order to get the main model parameters, the

derivation of the costing function J was achieved as

it is described by the following equation, in which θ

is the physical parameters matrix, W is the weights

matrix, Y and Γ are composed on the real and

imaginary coefficients from the linear regression

model:

It means, that parameters are shown in equation 7

that depends on the adaptive coefficients:

On the other hand, it is obtained the estimated

response Ŷ:

In which, due to get the adaptive estimation (it is

looking for optimal predictions), it is necessary the

weight matrix W:

W and W depend on adjusted coefficients ρ that is

C m

inside the range between -1 and 1 (values were

chosen for W and Wrespectively).

C m

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Predictive algorithm or optimal preventions of COVID-19

Hence, the expected solution Ŷ is adapted to its own

dynamic that was observing the dynamic of

another countries too, which give solution ranges

for W. It means, whether an expected curve P (as it

is depicted in the following figure) needs the

prediction to continue the trajectory from the point

Q, this trajectory can follow 3 possibilities by the

indicated arrows (up, down or right side).

Notwithstanding, it is suggested in this research to

see this problem as a complex statistic task and this

is the reason, why it is analysed as a dynamic

system, in which the prediction needs a reference

trajectory owing to get this road. Moreover, it is

necessary to recognize that every curve has its own

dynamic (O, G, B) with own difficulty to be

predicted (such as in the context of COVID-19), it

is suggested that adaptation factors need to be

correlated with adaptation factors of other

dynamics (from O, G, B) due to prevent or achieve

a better successful prediction of the new trajectory

of P. Despite, the road of P could change suddenly

(Bouchnita & Jebrane, 2020; De Falco et al., 2020;

Fang et al., 2020).

Figure 2. Dynamic curves for the proposed analysis.

The complexity of this proposal is given in changes

of the input matrix variables due to these excitation

signals take the information of the real response or

the expected response. It means for COVID-19

application (quantity of infected and deceased

people prediction) cannot achieve the desired

prediction, while it is not concentrated in the real

dynamic of the system (population changes) and to

make comparisons with other dynamics, some

curves just crossed the increasing transient and

achieved the decreasing response. What does this

context mean in the dynamic that was correlated for

the other curves? This was explained by the

compromise of the adaptation matrix of every

dynamic and the expected response as the

dependence on the parameters of the curves that got

a decreasing response. In this context, this is an

advantage of this proposal, because of COVID-19

context. It can be interpreted as every curve

represents countries dynamics with their own

coefficients and adaptations, which need to be

analysed for the dynamic of the system, which is

looking for its own prediction. That matrix can be

composed by this question: What medicine or

procedure support were prepared by the

government of countries where confinement only

cannot guarantee decreasing of the curve?

Furthermore, to get understanding of the behaviour

of every dynamic, it means there are countries with

more flexible rules to get a collective response, but

the flexibilities or other as a collective are not fast.

Therefore, there are countries with similar

dynamics if to compare with other. Hence, it can be

obtained matrixes and coefficients of an expected

dynamic.

However, what does it mean as a suggested answer

for the collective group? That translation is the

complex and it needs much communication

between authorities of different countries that

analysed countries dynamics, because perhaps

these coefficients can be translated as the medical

treatment between patients and medical doctors.

What support by medicine and what is the

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Calderón-Chavarri et al.

psychological answer for the collective group?

What is about the economical proposal solutions

for the collective? As it is proposed in this research,

every variable needs to be studied and correlated

and the result of the dynamics curves needs to be

interpreted according to get better prediction in

stochastic context. It is faster while medicine

answer accelerates this process (for example,

plasma analysis or increasing immunity of people).

To sum up, the data interpretation needs to be

analysed as the dynamic, but not to be isolated from

others, ever proposal solution needs to be adapted

for every reality.

RESULTS

Can quarantine days to achieve decreasing of the

quantity of infected people curve?

This question needs to be solved through optimal

predictions from the database that was published

by MINSA. In this context, it was modelled a

polynomial model algorithm, which is depicted in

the flowchart of Figure

Figure 3. Flowchart to describe the algorithm that was designed to predict the quantity of infected people during quarantine days.

The first step is given by the polynomial

identification from the input data. For this task it

was organized as input variables the quantity of

infected people, current days and quarantine days.

After it was necessary to calculate the matrix

weight by interruption of the main algorithm,

owing to adapt the weight coefficients. Finally, it

was analysed the error compromise by optimizing

the predicted quantity of infected people with the

expected response from its dynamic model that

was obtained before (from the polynomial model).

Hence, the algorithm Least Mean Square as

subroutine was the tester of the error of the result.

Executing the algorithm, it was possible to

understand the problem: the number of infected

people continues to grow as it is shown by the red

colour curve of Figure 4.

For this moment, in which it was achieved the last

result of the algorithm execution, it is March 19,

2020. It means the letter E of Figure 4 depicts the

quantity of infected people as 234, which is 89

more than the value in letter D (one day ago).

Therefore, from the red curve is showed that the

growth speed will continue until 440 infected

people in the eighth quarantine day with the error of

±3, only whether data published that was registered

by MINSA that keeps accuracy during the patients'

tests.

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Predictive algorithm or optimal preventions of COVID-19

The information that is showed in Figure 4 was

achieved during the middle of March 2020 (when

in Peru there was not registered deceased people

because of COVID-19). Figures 4 and 5 were

obtained as a consequence of registered data from

MINSA and the prediction result from the designed

algorithm (until April 6, 2020). In figure 5 is

depicted three internal figures due to analysing the

real data that was published by MINSA, even

though the real quantity of infected people is only

from the people, who made the test (whether

achieved infection or not). Therefore, to know the

real quantity of infected people in the country is not

simple, because the virus transmission can be given

from people, who are asymptomatic or who did not

make the infection test. However, for the analysis

to proportionate in this research, it is used official

data that was registered for MINSA.

In Figure 5A, it is depicted the curve (blue colour)

of virus infection evaluated people (quantity) E and

the curve (yellow colour) of no infected people,

both curves are under the dependence of the total

current days and quarantine days. It is possible to

see that during last days (since the 10-quarantine

day or the 25 current days) the quantity of non-

infected people started to decrease more and more

(more infected people were achieved).

In Figure 5B, it is showing the curve (green colour)

of the quantity of infected people I, the curve

(violet colour) - no deceased people quantity Nd

and the curve (red colour) – the quantity of

deceased people. It is possible to see that quantity

of deceased people started to increase since the 3rd

quarantine day.

In Figure 5C, it is showing the curve (red colour) of

deceased people (expanded view from Figure 5B).

Figure 4. The curve of the quantity of infected people as the dependence on current days and quarantine days that were joined to

the expected curve.

Figure 5. Curves of evaluated people, infected people, and deceased people for Peru, that were based on published data by

MINSA.

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Calderón-Chavarri et al.

In context to analyse an optimal prediction of the

variables: the quantity of infected people and the

quantity of deceased people, it was necessary to

give as the reference information of the designed

algorithm: the dynamic answer of China, Russia,

Spain and Italy to deal with COVID-19 (this

information is official and it was published by

Health Ministry of these countries). In China it was

reduced the number of infected people and the

quantity of deceased people before the 100 day

after infection started in this country. Therefore, the

red colour curve in Figure 6 depicts dynamic of

Peru, the blue colour curve depicts dynamic of

Russia, the yellow colour curve depicts dynamic of

Spain, the green colour curve depicts dynamic of

Italy.

Figure 6. The curve of the quantity of infected people as the dependence on current days and quarantine days that were joined to

the expected curve (Total stored amount).

This information was a reference variable to

achieve the Dynamic Peruvian model (red colour

curve) as it is showed in Figure 6. It was possible to

get the prediction of the maximal quantity of

infected people for every country from every letter

indicated that was indicated in Figure 6 (R, S, P, I).

It was correlated every curve that gives priority to

the coefficient weights that was obtained from the

countries data, which crossed the peak of their

curves (such as China, Spain, and Italy). Therefore,

the correlation of that dynamics in the main model

helped to finish the estimation for all of them and to

propose an optimal road for Spain, Italy, Russia and

Peru.

The blue colour rhombus represents the number of

the infected people since infection started in Russia

until the infection day number 103, in which was

achieved 485253 infected people approximately

(the rhombus R). After to correlate this curve with

the curves of other countries (Peru, Spain and

Italy), it was obtained the estimation that was

depicted by the blue colour curve (crossing

1400000 infected people until the day number

300). The red colour rhombus represents the

number of the infected people since infection

started in Peru until the infection day number 98, in

which was achieved 203736 infected people

approximately (the rhombus P). After to correlate

this curve with the curves of other countries

(Russia, Spain and Italy), it was obtained the

estimation that was depicted by the red colour

dotted curve (around 700000 infected people until

the day number 300, while it is not stopped the

quarantine days). The yellow colour rhombus

represents the number of the infected people since

infection started in Spain until the infection day

number 109, in which was achieved 241966

infected people approximately (the rhombus S).

After to correlate this curve with the curves of other

countries (Russia, Peru and Italy), it was obtained

the estimation that was depicted by the yellow

colour curve (around 295000 infected people until

the day number 300). The green colour rhombus

represents the number of the infected people since

infection started in Italy until the infection day

number 133, in which was achieved 235561

infected people approximately (the rhombus I).

After to correlate this curve with the curves of other

countries (Russia, Peru and Spain), it was obtained

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Predictive algorithm or optimal preventions of COVID-19

the estimation that was depicted by the green

colour curve (around 300000 infected people until

the day number 300).

Notwithstanding, the quarantine time was stopped

in Peru after 107 days and this effect caused that red

colour dotted curve changed its dynamic to the

expected red colour curve (showed in Figure 6).

The reason, why the number of infected people

increased, could be explained by social phenomena

that was correlated with economic problems for the

country. Therefore, the designed algorithm

predicted this change by the result of the red colour

curve.

By another side, there is another high quite

important variable to keep in mind that is the

quantity of deceased people as it is shown in Figure

7: the red colour is for Peru, the blue colour is for

Russia, the green colour is for Italy, the yellow

colour is for Spain.

Figure 7. The curve of the quantity of deceased people as the dependence on current days and quarantine days that were joined to

the expected curve.

The estimation can give the information to achieve

steady-state of the quantity of infected people and

deceased people, more or less since the 100 days of

the current days. Notwithstanding, it must to be

clear that this prediction could get good accuracy

only whether Peruvian government takes in control

variables as Chinese government did (to deal with

the virus). Italy, Spain and Russia are doing this (to

deal with the virus) as for example, to store people

inside homes in a strict way and medical

parameters for patients and population. But the

dynamic of Peru is not the same as the dynamic of

China due to many variables based in the economy,

population, behaviour of people etc. That is the

reason, why it is very important to study dynamics

from other countries too, in China it was achieved

good result and there are countries that cannot get it

yet. This is important to model the algorithm that

was designed according to be prepared for different

possibilities during next days, because to not make

mistakes to care life of humans, even though to

keep Peruvian population inside homes is

dangerous for economic variables that means to

correlate them in this research it needs next step.

The designed mathematical model takes as the

reference variables the dynamic models from other

countries owing to achieve adaptive coefficients

matrices, which can reduce the error as the

dependence of the main target: reduce virus

transmission and quantity of deceased people. The

blue colour rhombus represents the number of the

deceased people since the infection started in

Russia until the day number 103. Until that day,

approximately 6142 infected people died (the

rhombus R) after to correlate this curve with the

curves of other countries (Italy, Peru and Spain), it

was obtained the estimation that was depicted by

the blue colour curve (around 38000 died people

until the day number 300).

The red colour rhombus represents the number of

the deceased people since the infection started in

Peru until the day number 98. Until that day,

approximately 5738 infected people died (the

rhombus P) after to correlate this curve with the

curves of other countries (Italy, Russia and Spain),

it was obtained the estimation that was depicted by

the red colour curve (around 35000 died people

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Calderón-Chavarri et al.

until the day number 300).

The yellow colour rhombus represents the number

of the deceased people since the infection started in

Spain until the day number 109. Until that day,

approximately 27136 infected people died (the

rhombus S) after to correlate this curve with the

curves of other countries (Italy, Russia and Peru), it

was obtained the estimation that was depicted by

the yellow colour curve (around 29000 died people

until the day number 300).

The green colour rhombus represents the number

of the deceased people since the infection started in

Italy until the day number 132. Until that day,

approximately 33964 infected people died (the

rhombus S) after to correlate this curve with the

curves of other countries (Spain, Russia and Peru),

it was obtained the estimation that as depicted by

the green colour curve (around 37000 died people

until the day number 300).

It is necessary to analyse that the designed

algorithm estimated the effect of the increasing

number of infected people as the consequence of

the number of deceased people. Nevertheless, it is

predicted that it cannot be a significant cause to

make changes as it is shown in Figure 7.

It is proposed to study the dynamic of COVID-19

in every country as a polynomial model that can be

correlated each other according to optimize

prediction answers, there is no mathematical model

that adapts to the predictions such as the number of

infected people and the quantity of deceased people

(Kruse & Alkhushayni, 2020; Kucharski et al.,

2020; Marmarelis, 2020; Ping, 2020; Rustam et al.,

2020; Trigger & Czerniawski, 2020; Wang et al.,

2020; Yuankang et al., 2020). Hence, every

adaptive coefficients of matrix can be translated as

interchanging information between every country,

regarding to share many details and steps that could

deal with this pandemic. A good solution could be

to impose radical procedures to the population as

the dependence on researching immunity of

people. However, humans are not predictive

variables due to tests statistical methodologies that

is the reason, why in this research it is proposed

DISCUSSION

interchanging solutions between countries owing

to this pandemic is a world task now.

It is suggested to increase analysis by many

variables, such as the number of infected people by

different regions in Peru. Moreover, it is necessary

to compare curves with other countries results,

which crossed COVID-19 or which are under this

task too. The effect of external variables

(population parameters) helps to foresee an

economical effect in Peru, owing to the correlation

among the economy and the health can warrant

successful control in this epidemic situation.

It is expected to expand this research according to

study the consequences of this pandemic, through

correlating dynamics in economy, public health

and education from many countries around the

world dealing with this virus.

It is dedicated special gratitude to Hugo Medina,

because of his teachings in Science Physics for

many different generations of engineers, he did and

he makes that physics laws could be so easy to get

understanding of nature and current life, such as for

this research. With a very good base of laws of

nature, it was possible to obtain a fundament to

correlate advanced mathematics with the

formalism that engineering applications always

need. Even though Peru was not prepared to be face

to face against a big epidemic, but it was found the

answer from many researchers (read bellow), who

supported with their points of view, suggestions

and analysis discussions to finish this research,

which is waiting to be useful for the responsible

people, who have the task to organize priority of

activities. Because humans need to return to solve

tasks with much attention in physical parameters

and with caring much distance separations, room

temperature, room humidity, airflow and airspeed

between them. Therefore, it is expressed deep

warm thanks to Willy Gamboa, Christian Gozar,

Broni Huamaní, Alexánder Zutta, Leslie Vargas

and Lilian Gamarra, because of their suggestions

and opinions to analyze the consequences of this

research. It is expressed thankfulness to medical

doctors from the Health Center of PUCP, because

of their time to give suggestions in the development

ACKNOWLEDGMENT

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Predictive algorithm or optimal preventions of COVID-19

of this research. It is expressed thankfulness to

students of the lecture Nanotechnology MTR609,

Mechatronic Master Program, PUCP: G. Alvarado,

N. Azambuja, M. Chávez, I. Loayza, and M.J.

Romero, because of their opinions and suggestions

to analyse the consequence of this research.

BIBLIOGRAPHIC REFERENCES

Bouchnita, A. & Jebrane, A. 2020. A hybrid multi-

scale model of COVID 19 transmission

dynamics to assess the potential of non-

pharmaceutical interventions. Chaos,

Solitons & Fractals, 138: 109941.

Chen, Y.; Lu, P.; Chang, C. & Liu, T. 2020. A Time-

dependent SIR model for COVID-19 with

undetectable infected persons. IEEE

Transactions on Network Science and

E n g i n e e r i n g , 1 : d o i :

10.1109/TNSE.2020.3024723

Chowell, G.; Sattenspiel, L; Bansal, S. & Viboud,

C. 2016. Mathematical models to

characterize early epidemic growth: A

Review. Physics of Life Reviews, 18: 66-

97.

Coronavirus en el Perú. (2020,03,06). Diario El

C o m e r c i o . R e t r i e v e d f r o m

https://elcomercio.pe/peru/coronavirus-en-

peru-martin-vizcarra-confirma-primer-

caso-del-covid-19-en-el-pais-nndc-

noticia/?ref=ecr

De Falco, I.; Della-Cioppa, A.; Scafuri, U. &

Tarantino, E. 2020. Coronavirus COVID-19

spreading in Italy: optimizing an

epidemiological model with dynamic social

distancing through Differential Evolution.

R e t r i e v e d f r o m

http://arxiv.org/abs/2004.00553v3.

El Peruano. 2020. D. S. No 044-2020-PCM. Que

declara Estado de Emergencia Nacional por

las graves circunstancias que afectan la vida

de la Nación a consecuencia del brote del

COVID-19 Diario oficial del bicentenario

El Peruano (2020). Retrieved from

https://busquedas.elperuano.pe/normaslega

les/decreto-supremo-que-declara-estado-

de-emergencia-nacional-po-decreto-

supremo-n-044-2020-pcm-1864948-2/

Fang, Z.; Huang, Z.; Li, X.; Zhang, J.; Lv, W.;

Zhuang, L.; Xu, X & Huang, N. 2020. How

many infections of COVID-19 there will be

in the “Diamond Princess”-Predicted by a

virus transmission model based on the

simulation of crowd flow. Retrieved from

http://arxiv.org/abs/2002.10616

Kruse, R. & Alkhushayni, S. 2020. Identifying

regional COVID-19 presence early with

time series analysis. IOP SciNotes, 1:

024003.

Kucharski, A.; Russell, T.; Diamond, C.; Liu, Y.;

Edmunds, J.; Funk, S. & Eggo, R. 2020.

Early dynamics of transmission and control

of COVID-19: a mathematical modelling

study. The Lancet: Infectious diseases, 20:

553-558.

Marmarelis, V. 2020. Predictive modeling of

Covid-19 data in the US: Adaptive phase-

space approach. IEEE Open Journal of

Engineering in Medicine and Biology, 1:

doi: 10.1109/OJEMB.2020.3008313.

Pearson, A. 1979. Nonlinear system identification

with limited time data. Automatica, 15: 73-

84.

Ping, H. 2020. Study on epidemic prevention and

control strategy of COVID -19 based on

pe r so nn e l flo w pre d ic ti o n. 20 20

International Conference on Urban

Engineering and Management Science

(ICUEMS), 1: 688-691.

Rustam, F.; Reshi, A.; Mehmood, A.; Ullah, S.; On,

B.; Aslam, W. & Choi, G. 2020. COVID-19

Future Forecasting Using Supervised

Machine Learning Models. IEEE Access, 8:

101489-101499.

Trigger, S. & Czerniawski, E. 2020. Equation for

epidemic spread with the quarantine

measures: application to COVID-19.

Physica Scripta, 95:105001.

Wang, Z.; Shi, Y. & Xie, J. 2020. Application of

Logistic Model in COVID-19. Journal of

Physics: Conference Series, 1634: 012049.

Yuankang, Z.; Yi, H. & Xiaosong, Z. 2020.

COVID-19 Outbreak Prediction Based on

SEIQR Model. 2020 39 Chinese Control

Conference (CCC), 1: 1133-1137.

Received July 22, 2020.

Accepted December 31, 2020.

The Biologist (Lima). Vol. 19, Nº1, jan - jun 2021

Calderón-Chavarri et al.