ArticlePDF Available

The Metabolism Grey Prediction Model Based on Big Data and Internet of Things Technology


Abstract and Figures

In view of the uncertainty and diversity of the metabolic grey prediction model in the prediction process, resulting in poor prediction effect, a metabolic grey prediction model based on big data and Internet of things technology is constructed. Establish the aerobic metabolic process of human telecontrol and put forward the detection index of aerobic metabolic cycle function in human telecontrol; on this basis, use the metabolic grey prediction model analysis algorithm to determine the active intrusion intention of complex network, establish the intrusion intention attack behavior set function, establish the internal operation architecture under the technology of big data and Internet of things, and realize the construction of metabolic grey prediction model. The experimental results show that the constructed model can realize data prediction, with high confidence level and good effect.
This content is subject to copyright. Terms and conditions apply.
Research Article
The Metabolism Grey Prediction Model Based on Big Data and
Internet of Things Technology
Rong Fan, Suqin Sun, Yuanli Shan, Qingjiang Zhang, Jiazhuo Li, Guoqing Ma,
and Limin Pan
Heilongjiang University of Traditional Chinese Medicine, Harbin 150080, China
Correspondence should be addressed to Limin Pan;
Received 7 March 2022; Revised 24 March 2022; Accepted 30 March 2022; Published 14 April 2022
Academic Editor: Zhiguo Qu
Copyright © 2022 Rong Fan et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
In view of the uncertainty and diversity of the metabolic grey prediction model in the prediction process, resulting in poor
prediction eect, a metabolic grey prediction model based on big data and Internet of things technology is constructed.
Establish the aerobic metabolic process of human telecontrol and put forward the detection index of aerobic metabolic cycle
function in human telecontrol; on this basis, use the metabolic grey prediction model analysis algorithm to determine the
active intrusion intention of complex network, establish the intrusion intention attack behavior set function, establish the
internal operation architecture under the technology of big data and Internet of things, and realize the construction of
metabolic grey prediction model. The experimental results show that the constructed model can realize data prediction, with
high condence level and good eect.
1. Introduction
Metabolism includes material metabolism and energy
metabolism. It is composed of two opposite and simulta-
neous processes: assimilation and alienation. Assimilation
and alienation have obvious dierences and close relations.
Without assimilation, organisms cannot produce new proto-
plasm and store energy, and alienation cannot be carried
out. On the contrary, if there is no alienation, there can be
no energy release, and the material synthesis in organisms
cannot be carried out. It can be seen that assimilation and
alienation are both opposite and unied, which jointly deter-
mine the existence and continuity of organisms. In the long-
term evolution process, organisms constantly interact with
their environment and gradually form dierent types in
the way of metabolism. Therefore, assimilation is a process
of absorbing energy. For example, green plants use photo-
synthesis to convert water, carbon dioxide, and other sub-
stances in the environment into starch, cellulose, and other
substances. On the contrary is alienation, that is, from the
body to the external environment, the process of substances
from macromolecules to small molecules is a process of
releasing energy and expelling substances that organisms
do not need or cannot use. The grey metabolism model is
a constantly consider new information prediction model, it
considers the over time one after another into the system,
the eects of disturbance factors, at the same time of con-
stantly added new information, get rid of old information
in time, and make the whole system updated and the devel-
opment process, more in line with the change of real world
[1]. It can not only give full play to the advantages of tradi-
tional prediction models that only use a small amount of
data [2, 3] but also reect the changing trend of data, so as
to further improve the accuracy of prediction results.
Big data refers to the data set that cannot be captured,
managed, and processed by conventional software tools
within a certain time range. It is a massive, high-growth,
and diversied information asset that requires a new pro-
cessing mode to have stronger decision-making ability,
insight, and discovery ability and process optimization abil-
ity [4, 5]. Big data has 5 V characteristics: volume, velocity,
variety, value, and veracity. Big data itself is an abstract con-
cept [68]. Relying on the development of the Internet and
cloud computing, big data plays an increasingly important
Wireless Communications and Mobile Computing
Volume 2022, Article ID 6106995, 9 pages
role in all walks of life [9]. The Internet of Things is the
Internet of everything, namely, the Mobile Internet, which
is no longer the link between people but machine to
machine, device to device, and system to system [10]. In this
regard, reference [11] proposed using big data to guide bet-
ter nurse allocation strategies, using machine learning
method to predict hospital discharge, discrete event simula-
tion model to determine the needs of nurses in neonatal
ICU, and using machine learning and hierarchical linear
regression to connect the changing nurse allocation with
the results of patients. This new study applied a unique
Monte Carlo simulation model to estimate nursing needs
and test dierent strategies to meet needs. Reference [12]
proposed the mortality, disability, and recurrence rate after
the rst stroke, which is a study from the Chinese stroke
big data observation platform. Stroke mortality, disability,
and recurrence rates were investigated 12 months after the
rst stroke, a prospective cohort study based on national
hospitals. The intravenous thrombolysis rate was 9.5%, and
the intravascular treatment rate was 4.4%. The results sup-
port the hypothesis that the prognosis of stroke patients in
China seems to have improved, not very bad.
Although the above research has made some progress, it
is not suitable for newborn metabolism prediction. There-
fore, a metabolism grey prediction model based on big data
and Internet of things technology is proposed. Big data can
be understood as a resource or asset. With the increasingly
powerful processing capacity of computers, the more data
can be obtained, and the more value can be mined. The
Internet of Things is also the most critical link in an intelli-
gent society, but it is often ignored by people. As the termi-
nal closest to users, it is not only the front line for collecting
large-scale and multidimensional user data but also the win-
dow and bridge for the eects of technology to be feedback
to user experience. Big data and the Internet of things sound
like them knowing you cannot have them, very tallsmart
technology, but it is closely related to many aspects of our
social life, and it works from the institute to the life in every
scene, related to the business analysis, data analysis, data
mining, machine learning, articial intelligence, and other
elds, from pure technical research to specic application.
Smart technology is everywhere. Using big data and Internet
of Things technology to build a grey prediction model of
metabolism has good performance.
2. Analysis of Aerobic Metabolic Function of
Human Exercise
2.1. Human Telecontrol Aerobic Metabolic Process. The pro-
cess of human telecontrol is mainly aerobic metabolism.
Aerobic metabolic exercise refers to the endurance exercise
and exercise for the purpose of enhancing human oxygen
inhalation, oxygen transmission, and oxygen metabolism.
The benets of aerobic metabolic exercise are as follows: it
can gradually increase the amount of oxygen absorbed by
the human body during exercise. In the metabolic process,
it can better consume the excess heat in the human body,
so as to metabolize the excess heat in the body. In other
words, in the process of human telecontrol, the amount of
oxygen inhaled by the human body is basically equal to the
amount of oxygen required, so as to achieve physiological
metabolic balance in the process of circulation. Therefore,
aerobic metabolic exercise is characterized by rhythm, unin-
terrupted, and long duration. Predict the data transmission
according to the data transmission mode of the rule base.
Construct the human telecontrol aerobic metabolism
scheme and set the metabolic structure, as shown in
Figure 1.
According to Figure 1, in the process of human telecon-
trol, aerobic metabolism is mainly the metabolism of glu-
cose. The glucose stored in human body forms 6-P glucose
through metabolic reaction with ATP and then reacts with
NADP to form 6-P gluconic acid. After dehydration, it
forms the intermediate 2-keto-3-deoxy-6-phosphogluconic
acid and nally forms pyruvate ×2.
Pyruvate, as a main product of metabolism, can form the
energy required by human functional exercise through sub-
sequent aerobic metabolism. Through the metabolic reaction
with a variety of human substances, the energy required for
human movement is nally formed. The key to aerobic met-
abolic exercise is to master the appropriate amount of exer-
cise, which requires not only certain competition and
exercise intensity but also continuous metabolic ability.
2.2. Detection Index of Aerobic Metabolism and Circulation
Function in Human Telecontrol. Based on the above analysis
of aerobic metabolism process in human telecontrol,
through comprehensive modeling and analysis of various
factors aecting aerobic cycle metabolism in human telecon-
trol, a dynamic balance model of aerobic metabolism and
supply is constructed to provide guidance for aerobic metab-
olism analysis of human telecontrol.
Based on the analysis model, the parameters aecting
aerobic exercise are analyzed. When analyzing the function
of human telecontrol aerobic metabolic cycle, taking 10
domestic elite athletes as the object, through long-time team
competition, physical index test, training intensity, and time
analysis, the physiological index test database is constructed
as the research data sample of this paper. It is assumed that
there is a prediction data in the data sample, and the data is
located in a level. Before the assumption of information data,
the standard processing method is adopted to strengthen the
processing of the assumption mode, and the scheme is for-
mulated according to the set human telecontrol aerobic
metabolism structure. The execution structure is shown in
Figure 2.
According to Figure 2, adjust the structural tting angle
of demand prediction, control the demand of aerobic meta-
bolic cycle function according to the angle information,
check the consistency of the analyzed data, store the tested
data in the model construction system, and wait for the sub-
sequent model design operation.
In the process of team training and competition moni-
toring, parameter comparison research is mainly carried
out on the athletesshortest aerobic metabolism time, aero-
bic metabolism eciency, and aerobic metabolism ratio, in
order to have a good measurement and calibration for the
athletesmetabolic process and metabolic ability. In the
2 Wireless Communications and Mobile Computing
process of aerobic metabolism, various parameters mainly
have the following functions:
(1) The shortest aerobic metabolic time: it mainly
reects the time for athletes to change from anaero-
bic metabolic state to aerobic metabolic state and
reects the adaptability of athletes to the
(2) Eciency of aerobic metabolism: it reects the e-
ciency of dierent athletesbodies in converting sub-
stances into energy under the condition of the same
amount of oxygen inhalation. The higher the e-
ciency, the stronger the ability of metabolism
(3) Aerobic metabolism ratio: in the whole metabolic
process, the ratio between aerobic metabolism and
anaerobic metabolism reects the composition and
distribution of athletesmetabolism
Based on the detailed test of the above indexes, the met-
abolic grey prediction model is constructed, and the aerobic
metabolic cycle function in human telecontrol is analyzed
through the evaluation of functional index. Through the
above analysis, it can be seen that in the process of human
telecontrol, with the process of long-distance running,
irregular attenuation of respiration occurs, and the process
of aerobic exercise is mixed with irregular anaerobic
2.3. Metabolic Grey Prediction Model Analysis Algorithm.
Through the construction of the above human telecontrol
aerobic metabolic function and evaluation parameter index
system, the original sample data of athletesaerobic meta-
bolic function analysis and evaluation index [13, 14] are
obtained. Based on the analysis method of metabolic grey
prediction model, the athletesaerobic metabolic cycle func-
tion is analyzed, and the decline characteristics of the model
are used to characterize the aerobic metabolic consumption.
Taking the shortest aerobic metabolism time, aerobic metab-
olism eciency, and aerobic metabolism ratio as the model
input parameters, the model input matrix of three parame-
ters is constructed as follows:
Kmn =
k11 k12 k13
k21 k22 k23
k31 k32 k33
In formula (1), k11 represents the autocorrelation charac-
teristic of the shortest aerobic metabolism time, k22 repre-
sents the autocorrelation characteristic of the eective rate
of aerobic metabolism, and k33 represents the autocorrela-
tion characteristic of the rate of aerobic metabolism.
The autocorrelation characteristics of the input parame-
ters of the model characterize the stability of the model.
Based on the construction of the input parameter matrix,
the metabolic grey prediction model equation is obtained
as follows:
Abc =
0:85 0:85/z10:85/z2
1:00 1:00/z11:00/z2
0:95 0:95/z10:95/z2
In formula (2), Abc represents the characteristic equation
of metabolic grey prediction model, z1represents the corre-
lation distance between the shortest aerobic metabolic time
and aerobic metabolic eciency, and z2represents the corre-
lation distance between the shortest aerobic metabolic time
and aerobic metabolic ratio. The analysis process of meta-
bolic grey prediction model is as follows:
Step 1. Initialize Ddd:
Ddd =
d11 d12 d13
d21 d22 d23
d31 d32 d33
In formula (3), Ddd represents the fading factor matrix of
the metabolic grey prediction model. The three autocorrela-
tion characteristics in the matrix correspond to the parame-
ter fading autocorrelation characteristics of the shortest
aerobic metabolism time, aerobic metabolism eciency,
and aerobic metabolism ratio.
Step 2. Calculate the parameter matrix Min the grey
prediction model of metabolism:
M=Ddd ×Abc:ð4Þ
Step 3. In the metabolic grey prediction model, move
forward along the same column according to the matrix dis-
tribution characteristics of the model [15, 16].
Step 4. Move to an adjacent column and calculate the
fading characteristic factor of the metabolic grey prediction
model, expressed as
RER =S×M×θ:ð5Þ
Case reuse
Figure 1: Structure of human telecontrol aerobic metabolism.
3Wireless Communications and Mobile Computing
In formula (5), Srepresents the fading factor of the met-
abolic grey prediction model, and θrepresents the fading
parameter distribution domain of the metabolic grey predic-
tion model;
Step 5, According to the model fading factor calculated
above, move to two adjacent columns. After the new value
in the model is updated, the new value represents the aerobic
metabolic consumption of the current athlete [17].
Step 6, Calculate the aerobic metabolic recovery of ath-
letes according to the recovery characteristics of fading fac-
tors and get
In formula (6), λrepresents aerobic metabolic function
index, that is, the output parameter of metabolic grey predic-
tion model.
To sum up, by constructing the metabolic decline char-
acteristics, take the shortest aerobic metabolism time, aero-
bic metabolism eciency, and aerobic metabolism ratio as
the input parameters of the metabolic grey prediction model,
take the overall aerobic metabolic function index as the
model output parameters, represent the aerobic metabolic
consumption through the decline characteristics, and reect
the nal aerobic metabolic balance through the aerobic met-
abolic function index.
3. Realize the Metabolic Grey Prediction Model
Based on Big Data and Internet of
Things Technology
3.1. Internal Operation Architecture under Big Data and
Internet of Things Technology. Under the technology of big
data and Internet of things, in the process of metabolic pre-
diction of human function, rst determine the active intru-
sion intention of complex network [1820], establish the
set function of intrusion intention and attack behavior, and
estimate the diusion equation of active intrusion metabo-
lism of complex network on this basis. Based on the diu-
sion equation, a grey prediction model of active intrusion
metabolism in complex networks is constructed [21, 22].
Assuming that the metabolic diusion feature extracted
under the active intrusion of complex network is α, the state
of metabolic diusion under the active intrusion of complex
network is
In formula (7), Mirepresents the weight coecient of
each metabolic diusion feature, and conαrepresents the
observation sequence of metabolic diusion state under
complex network intrusion [23], which needs to meet the
conditions of i=1,2,3,4.
According to the conclusions drawn from the above pro-
cess, a grey prediction model of active intrusion metabolism
in complex networks is constructed [24, 25]. At the same
time, build the internal operations architecture, as shown
in Figure 3.
According to Figure 3, in the internal operation architec-
ture, the division of modules is not arbitrary, but should fol-
low a certain principle, and the internal correlation of
modules should be as close as possible. In this way, the
divided modules have a certain independence, so as to
reduce the complex calling relationship between modules
and make the structure of the operating system clear. The
internal parts of the module are closely linked, so that each
module has independent functions.
3.2. Construction of the Metabolic Grey Prediction Model. In
the process of constructing the metabolism grey prediction
model, based on the evaluation results of the complex net-
work active intrusion metabolism diusion process [26],
the state transition probability value of the complex network
active intrusion metabolism in the diusion process is calcu-
lated by means of probability reasoning [27, 28]. The basic
principle of time series prediction is based on the trend pre-
diction principle. The grey prediction theory is established
by using the methods of accumulationand subtraction.
When there is no obvious trend in the time series, the accu-
mulation method can be used to generate the time series
with obvious trend. According to the growth trend of the
series, the prediction model can be established, the inuence
Step 1
Step 2
Step 3
Output 1
Output 2
Formulation stage Implementation phase
Figure 2: Scheme implementation structure.
4 Wireless Communications and Mobile Computing
of grey factors can be considered for prediction, and then the
subtractionmethod is used for inverse operation to restore
the original time series. Based on the prediction results, a
grey prediction model of active intrusion metabolism in
complex networks is constructed.
Let the weight coecient between the j-th complex net-
work active intrusion metabolism diusion state and the i-th
diusion direction be Dji, and the connection parameter
between the jcomplex network active intrusion metabolism
diusion state and the idiusion direction be Eji .Dene the
diusion direction of the new generation of complex net-
work active intrusion as
In formula (8), Qjrepresents the initial state probability
distribution value in the metabolic diusion process of active
intrusion of complex network, and Pirepresents the meta-
bolic diusion random variable under active intrusion of
complex network.
Let the transfer function of metabolic diusion under
active intrusion of complex network be Mzx, and the calcu-
lation formula is
Mzx =l1
j=2 f2
In formula (9), lrepresents the number of metabolic dif-
fusion samples under complex network active intrusion, f2
represents the belief conversion function of metabolism
under complex network active intrusion, and grepresents
the direction of metabolic diusion under complex network
active intrusion [29, 30].
In the process of constructing the metabolic grey predic-
tion model of complex network active intrusion, it is
assumed that C1and C2represent the compensation coe-
cient of metabolic diusion prediction samples under com-
plex network active intrusion, xmax represents the
maximum empirical weight of metabolic diusion depen-
dence under complex network active intrusion, and xmin
represents the minimum empirical weight of metabolic dif-
fusion dependence under complex network active intrusion.
If the conditional probability change process of metabolic
diusion under active intrusion of complex network is stable
to the direction set potential in a limited time [31, 32], the
joint distribution probability of metabolic diusion of active
intrusion of complex network is obtained as follows:
Ggg =C1×C2
In formula (10), Iimax represents the observation func-
tion of metabolic diusion under active intrusion of complex
network. The objective function of metabolic diusion of
active intrusion of complex network is calculated by formula
(11). The formula is
Zmb = arg max Zt
In formula (11), ZðtÞrepresents the diusion direction
of complex network active intrusion metabolism at ttime,
parepresents the diusion state set of complex network
active intrusion metabolism, prepresents the diusion direc-
tion set of complex network active intrusion metabolism,
and Zðt1Þrepresents the new direction of complex net-
work active intrusion metabolism diusion state t1[33,
According to the derivation of the above process, a grey
prediction model of active intrusion metabolism in complex
networks is constructed, which is expressed as
MX= arg max pZ
To sum up, by determining the set function of active
intrusion intention and attack behavior of complex net-
works, the metabolic diusion process of active intrusion
of complex networks is estimated [35, 36]. Based on the eval-
uation results of metabolic diusion process of active intru-
sion of complex networks, a grey prediction model of active
intrusion metabolism of complex networks is constructed,
and the construction of metabolic grey prediction model
based on big data and Internet of things technology is
4. Experimental Analysis
In order to verify the eect and feasibility of metabolic grey
prediction model based on big data and Internet of things
technology, simulation experiments are set up. The coding
of the experiment is realized by MATLAB R2016a software
platform. The hardware environment adopts Intel Core i5-
3570 model 3.4GHz processor, and 8 GB installed memory,
and the operating system is 64 bit Windows7 agship. The
simulation tool for data information processing is imple-
mented in C++. 100 nodes are randomly deployed in the
grey prediction model to collect metabolic information.
The coverage of each sensor node is 0.93, the xed delay of
wireless sensor network routing transmission is 1.5 ms, the
output carrier freque ncy is 6.8 kHz, and the data size is
Structure 1
Structure 2
Data fetch
Data request
Send out
Figure 3: Internal operations architecture.
5Wireless Communications and Mobile Computing
1200 bytes. Other experimental parameters are set as shown
in Table 1.
Set the experimental scenario according to the parame-
ters in Table 1 and x the metabolic information within
the prediction range of the model to prevent experimental
errors caused by too small prediction range. At the same
time, strengthen the prediction management of metabolism,
adjust the allocation principle, execute the nal prediction
instruction, obtain the required prediction result data, and
extract the demand analysis performance data from the pre-
diction result data. The operation process of MATLAB soft-
ware is shown in Figure 4.
According to the parameter setting in Table 1 above and
the experimental operation ow in Figure 4, the metabolic
grey prediction model is veried, and dierent methods are
tested for metabolic topological grey prediction. Taking ref-
erence [11] and reference [12] as comparison methods, the
condence interval distribution of metabolic topological
grey prediction is obtained, as shown in Figure 5.
According to the analysis of Figure 5, the grey prediction
of metabolic topology is carried out by the model in this
paper. The condence level of data prediction realized by
sensor nodes is high and has good eect, while the predic-
tion eect of reference [11] and reference [12] is relatively
poor. The reason is that this model establishes the process
of human telecontrol aerobic metabolism, which can gradu-
ally increase the amount of oxygen absorbed by the human
body during exercise. In the process of metabolism, it can
better consume the excess heat in the human body and
metabolize the excess heat in the body, so as to enhance
the level of condence.
In order to further verify the eect of the model in this
paper, taking the training and competition data of 10 ath-
letes as the data basis for the analysis of metabolic grey pre-
diction model, the data test and result analysis were carried
out. The follow-up data statistics of half a year are adopted.
The interval of each test is no less than 2 hours, and each test
is carried out three times, aiming at the dierences caused by
gender and age. The correction factor method is used to cor-
rect the error caused by individual dierences, and the
detailed data distribution is shown in Table 2.
Based on the sample data in Table 2, the metabolism
grey prediction model function analysis experiment of aero-
bic metabolism function is carried out. From the experimen-
tal results and data analysis in Table 2, it can be concluded
that the shortest aerobic metabolism time of male athletes
is shorter than that of female athletes, mainly because men
store more energy; so, anaerobic exercise is more developed.
The aerobic metabolic rate varies greatly due to individual
dierences, which is the representation of the overall metab-
olism of athletes, through statistical analysis software SPSS
10.0. Dierent parameters are processed to obtain the cor-
rection factors under dierent individual dierences. The
results are shown in Table 3.
The unied results of aerobic metabolic function under
the grey prediction model are obtained through the correc-
tion of correction factors, as shown in Table 4.
The aerobic metabolic function of 10 athletes is analyzed
by metabolic grey prediction model, and multistep iteration
is carried out to obtain the error convergence curve of aero-
bic metabolic function analysis, as shown in Figure 6.
Table 1: Experimental parameters.
Project Parameter
Forecast data Material demand data
Database SCADA database
Characteristic factors Similarity parameter
Set value maximum demand Fifty thousand
Coecient adjustment Solar term coecient
Algorithm management Machine learning management
Prediction network Big data and internet of things technology
Open MATLAB soware
and import sample data
Training metabolic
information sample data
Load different grey prediction
models respectively
Test training set
Data normalization
processing and output
Figure 4: Operation ow of MATLAB software experiment.
6 Wireless Communications and Mobile Computing
It can be seen from Figure 6 that the error of aerobic
metabolic function analysis converges rapidly and has good
applicability to dierent individual dierences, indicating
that the constructed metabolic grey prediction model ana-
lyzes the characteristics of athletesaerobic metabolic func-
tion, and the model has good robustness.
To sum up, the accuracy of the metabolic function pre-
diction of athletes can be achieved based on the high reliabil-
ity of the Internet of things, and the reliability of the
metabolic function prediction based on grey technology is
good in the current application of the Internet of things.
5. Conclusions and Prospects
5.1. Conclusions
(1) The sensor nodes of the metabolism grey prediction
model designed based on big data and Internet of
Things technology achieve a high level of condence
in data prediction and have good eects
(2) The ratio of oxygen metabolism varies greatly with
individual dierences
(3) The metabolic grey prediction model was established
to analyze the aerobic metabolic function character-
istics of athletes, and the model had good robustness
5.2. Prospects
(1) The next step is to promote the feedback regulation
of human metabolism through aerobic exercise
training, so that more oxygen can be delivered to
brain cells in time. Aerobic exercise can improve
the level of human metabolism, eectively increase
the fat burning rate, ensure the eective operation
of body functions, and ensure that the body is in a
healthy state
(2) Due to the reduced ability of human body to secrete
antioxidants and enzymes, the level of free radicals
in the body increases. Lipid alcohols such as triglyc-
erides are easy to combine with free mineral ions in
5 1015202530
Transmission confidence
Time (s)
Paper model
Reference [11] model
Reference [12] model
Figure 5: Condence interval of metabolic topological grey
Table 2: Distribution of human telecontrol test data.
Athletes Gender
Minimum aerobic
metabolism time
eciency (%)
1 Female 0.15 90 105 : 1
2 Female 0.23 91 116 : 1
3 Female 0.19 94 123 : 1
4 Female 0.20 95 110 : 1
5 Female 0.18 91 115 : 1
6 Male 0.29 89 131 : 1
7 Male 0.30 93 108 : 1
8 Male 0.39 97 109 : 1
9 Male 0.29 89 131 : 1
10 Male 0.30 93 108 : 1
Table 3: Correction factors of dierent individual dierences.
Maximum heart
15-18 1.10 210 1.12
19-22 0.83 180 0.83
23-26 0.78 170 0.75
27-30 0.71 150 0.64
Table 4: Analysis results of the aerobic metabolism grey prediction
Athletes Oxygen absorption
Aerobic metabolic
function index
129~34 35~43 44~48
239~43 44~51 52~56
328~33 34~47 42~47
435~39 40~47 48~51
526~31 32~40 41~45
631~35 36~43 44~47
722~28 29~36 37~41
835~39 32~39 40~43
0 1020304050607080 90 100
Iteration steps/Frequency
Iterative error
Figure 6: Error convergence curve of aerobic metabolic function
7Wireless Communications and Mobile Computing
the body, forming diseases such as obesity and
hyperthyroidism. The future research content needs
to carry out fat burning through aerobic exercise to
promote human metabolism. In the process of aero-
bic exercise, skin blood vessels contract sharply, and
a large amount of blood is sucked into internal
organs and deep tissues to enhance human
(3) Carrying out aerobic metabolism training can
improve human endurance and impact explosive-
ness, guide scientic group training based on meta-
bolic grey prediction model, and promote the
coordinated development of human aerobic metabo-
lism and anaerobic metabolism, so as to improve the
level of physical health
Data Availability
The raw data supporting the conclusions of this article will
be made available by the authors, without undue reservation.
Conflicts of Interest
The authors declared that they have no conicts of interest
regarding this work.
This work supported by Rong Fan would like to acknowl-
edge the support of the National Natural Science Foundation
of Heilongjiang Province (No. H2018059) and Heilongjiang
Province Foundation for Returness.
[1] S. Bag, S. Gupta, T. Choi, and A. Kumar, Roles of innovation
leadership on using big data analytics to establish resilient
healthcare supply chains to combat the covid-19 pandemic: a
multi-methodological study,IEEE Transactions on Engineer-
ing Management, vol. 22, no. 8, pp. 114, 2021.
[2] W. J. Tu, B. H. Chao, L. Ma, F. Yan, and L. D. Wang, Case-
fatality, disability and recurrence rates after rst-ever stroke:
a study from bigdata observatory platform for stroke of
China,Brain Research Bulletin, vol. 175, no. 48, pp. 130
135, 2021.
[3] M. T. Odenkirk, D. M. Reif, and E. S. Baker, Multiomic big
data analysis challenges: increasing condence in the interpre-
tation of articial intelligence assessments,Analytical Chem-
istry, vol. 93, no. 22, pp. 77637773, 2021.
[4] D. Gathers, V. S. Pankratz, M. Kosich, and B. Tawk, Using
big data to gauge eectiveness of breast cancer awareness
month,Preventive Medicine, vol. 150, no. 8, article 106695,
[5] W. P. Tsai, D. Feng, M. Pan et al., From calibration to param-
eter learning: harnessing the scaling eects of big data in geos-
cientic modeling,Nature Communications, vol. 12, no. 1,
pp. 113, 2021.
[6] A. Bakdi, I. K. Glad, and E. Vanem, Testbed scenario design
exploiting trac big data for autonomous ship trials under
multiple conicts with collision/grounding risks and spatio-
temporal dependencies,IEEE Transactions on Intelligent
Transportation Systems, vol. 22, no. 12, pp. 79147930, 2021.
[7] N. Queiroz, N. E. Humphries, A. Couto, M. Vedor, and
D. W. Sims, Reply to: caution over the use of ecological
big data for conservation,Nature, vol. 595, no. 7866,
pp. E20E28, 2021.
[8] F. X. Goudot, S. Msadek, T. Boukertouta, P. O. Schischmano,
and C. Meune, Routine use of natriuretic peptides: lessons
from a big data analysis,Annals of Clinical Biochemistry,
vol. 58, no. 5, pp. 481486, 2021.
[9] M. A. Millis, S. L. Cramm, and M. Englesbe, Big data for small
öeople: how novel data collection can improve pediatric trans-
plantation outcomes,Liver Transplantation, vol. 28, no. 3,
pp. 361-362, 2022.
[10] B. Sliwa, R. Adam, and C. Wietfeld, Client-based intelligence
for resource ecient vehicular big data transfer in future 6g
networks,IEEE Transactions on Vehicular Technology,
vol. 70, no. 6, pp. 53325346, 2021.
[11] J. Spetz, Leveraging big data to guide better nurse stang
strategies,BMJ Quality and Safety, vol. 30, no. 1, pp. 97
103, 2020.
[12] O. Kulkarni, S. Jena, and V. R. Sankar, Mapreduce framework
based big data clustering using fractional integrated sparse
fuzzy c means algorithm,IET Image Processing, vol. 14,
no. 12, pp. 27192727, 2020.
[13] G. Tikhonov, L. Duan, N. Abrego et al., Computationally e-
cient joint species distribution modeling of big spatial data,
Ecology, vol. 101, no. 2, pp. e02929e02938, 2020.
[14] D. Sujatha, M. Subramaniam, and C. R. R. Robin, A new
design of multimedia big data retrieval enabled by deep feature
learning and adaptive semantic similarity function,Multime-
dia Systems, vol. 5, no. 2, pp. 1523, 2022.
[15] D. Fnyes, B. Nmeth, and P. Gspr, Design of lpv control for
autonomous vehicles using the contributions of big data anal-
ysis,International Journal of Control, vol. 8, no. 1, pp. 112,
[16] M. Poon, J. Gaete-Villegas, P. M. Brennan, and J. Fleuriot,
Letter to the editor. Misconceptions in the eld guide to big
data for neurosurgeons,Journal of Neurosurgery, vol. 135,
no. 2, pp. 112, 2021.
[17] R. Skuratovskii, A. Bazarna, and E. Osadhyy, Analysis of
speech mel scale and its classication as big data by param-
eterized knn,Articial Intelligence, vol. 26, no. 1, pp. 42
57, 2021.
[18] F. Aghaeipoor, M. M. Javidi, and A. Fernandez, Ifc-bd: an
interpretable fuzzy classier for boosting explainable articial
intelligence in big data,IEEE Transactions on Fuzzy Systems,
vol. 6, no. 99, pp. 18, 2021.
[19] J. C. Perafan-Villota, O. H. Mondragon, and W. M. Mayor-
Toro, Fast and precise: parallel processing of vehicle trac
videos using big data analytics,IEEE Transactions on Intelli-
gent Transportation Systems, vol. 9, no. 99, pp. 110, 2021.
[20] R. Ramoni, M. Klote, S. Muralidhar et al., Covid-19 insights
partnership: leveraging big data from the department of vet-
erans aairs and supercomputers at the department of energy
under the public health authority,Journal of the American
Medical Informatics Association, vol. 28, no. 7, pp. 1578
1581, 2021.
[21] J. T. Alaimo and C. J. Saunders, Big data strikes again: future
utilization of the Uk biobank as a resource for clinical labora-
tories,Clinical Chemistry, vol. 67, no. 7, pp. 932934, 2021.
8 Wireless Communications and Mobile Computing
[22] V. Kumar, D. K. Sharma, and V. K. Mishra, Mille cheval: a
gpu-based in-memory high-performance computing frame-
work for accelerated processing of big-data streams,The Jour-
nal of Supercomputing, vol. 77, no. 7, pp. 69366960, 2021.
[23] A. Aghdashi and S. L. Mirtaheri, Correction to: novel
dynamic load balancing algorithm for cloudbased big data
analytics,The Journal of Supercomputing, vol. 78, no. 3,
pp. 74797496, 2021.
[24] R. Susan and S. M. Purcell, Sleep and big data: harnessing
data, technology, and analytics for monitoring sleep and
improving diagnostics, prediction, and interventionsan era
for sleep-omics?,Sleep, vol. 44, no. 6, pp. 1523, 2021.
[25] T. Seki, M. Takeuchi, and K. Kawakami, Eating and drinking
habits and its association with obesity in japanese healthy
adults: retrospective longitudinal big data analysis using a
health check-up database,British Journal of Nutrition,
vol. 126, no. 10, pp. 15851591, 2021.
[26] P. Chawla, X. Cao, Y. Fu, C. M. Hu, and J. Z. Gao, Water qual-
ity prediction of salton sea using machine learning and big
data techniques,International Journal of Environmental Ana-
lytical Chemistry, vol. 8, no. 8, pp. 124, 2021.
[27] M. Zhang, J. Montewka, T. Manderbacka, P. Kujala, and
S. Hirdaris, A big data analytics method for the evaluation
of ship - ship collision risk reecting hydrometeorological con-
ditions,Reliability Engineering and System Safety, vol. 213,
no. 1, p. 107674, 2021.
[28] L. Morin and B. D. Onwuteaka-Philipsen, The promise of big
data for palliative and end-of-life care research,Palliative
Medicine, vol. 35, no. 9, pp. 16381640, 2021.
[29] H. Johnson, J. M. Davies, J. Leniz, E. Chukwusa, and K. E. Slee-
man, Opportunities for public involvement in big data
research in palliative and end-of-life care,Palliative Medicine,
vol. 35, no. 9, pp. 17241726, 2021.
[30] D. Mukherjee, Regarding "limitations in the analysis of ather-
ectomy using medicare big data",Journal of Endovascular
Therapy, vol. 28, no. 5, pp. 812-813, 2021.
[31] G. Carreño, A. Quesada, C. Rosado et al., Mo110are big-data
techniques useful in screening for primary hyperaldosteron-
ism? Results of a pilot study and their prospective application,
Nephrology, Dialysis, Transplantation,vol. 36, Supplement_1,
pp. 110, 2021.
[32] M. Than, S. Richardson, and J. Pickering, Emergency depart-
ment frequent attenders: big data insights for a big and com-
plex problem,Emergency Medicine Journal, vol. 39, no. 1,
pp. 2130, 2021.
[33] S. Mumtaz, A. Al-Dulaimi, H. Gacanin, and B. Ai, Block
chain and big data-enabled intelligent vehicular communica-
tion,IEEE Transactions on Intelligent Transportation Systems,
vol. 22, no. 7, pp. 39043906, 2021.
[34] X. Zhou, Dynamic decomposition simulation of spread Spec-
trum communication load based on Internet of Things tech-
nology. Computer,Simulation, vol. 37, no. 1, pp. 280283,
[35] D. Choi, H. Lee, K. Bok, and J. Yoo, Design and implementa-
tion of an academic expert system through big data analysis,
The Journal of Supercomputing, vol. 77, no. 7, pp. 78547878,
[36] F. M. Awaysheh, M. N. Aladwan, S. Alawadi, J. C. Cabaleiro,
and T. F. Pena, Security by design for big data frameworks
over cloud computing,IEEE Transactions on Engineering
Management, vol. 2, no. 99, pp. 118, 2021.
9Wireless Communications and Mobile Computing
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
Nowadays, multimedia big data have grown exponentially in diverse applications like social networks, transportation, health, and e-commerce, etc. Accessing preferred data in large-scale datasets needs efficient and sophisticated retrieval approaches. Multimedia big data consists of the most significant features with different types of data. Even though the multimedia supports various data formats with corresponding storage frameworks, similar semantic information is expressed by the multimedia. The overlap of semantic features is most efficient for theory and research related to semantic memory. Correspondingly, in recent years, deep multimodal hashing gets more attention owing to the efficient performance of huge-scale multimedia retrieval applications. On the other hand, the deep multimodal hashing has limited efforts for exploring the complex multilevel semantic structure. The main intention of this proposal is to develop enhanced deep multimedia big data retrieval with the Adaptive Semantic Similarity Function (A-SSF). The proposed model of this research covers several phases “(a) Data collection, (b) deep feature extraction, (c) semantic feature selection and (d) adaptive similarity function for retrieval. The two main processes of multimedia big data retrieval are training and testing. Once after collecting the dataset involved with video, text, images, and audio, the training phase starts. Here, the deep semantic feature extraction is performed by the Convolutional Neural Network (CNN), which is again subjected to the semantic feature selection process by the new hybrid algorithm termed Spider Monkey-Deer Hunting Optimization Algorithm (SM-DHOA). The final optimal semantic features are stored in the feature library. During testing, selected semantic features are added to the map-reduce framework in the Hadoop environment for handling the big data, thus ensuring the proper big data distribution. Here, the main contribution termed A-SSF is introduced to compute the correlation between the multimedia semantics of the testing data and training data, thus retrieving the data with minimum similarity. Extensive experiments on benchmark multimodal datasets demonstrate that the proposed method can outperform the state-of-the-art performance for all types of data.
Full-text available
The behaviors and skills of models in many geosciences (e.g., hydrology and ecosystem sciences) strongly depend on spatially-varying parameters that need calibration. A well-calibrated model can reasonably propagate information from observations to unobserved variables via model physics, but traditional calibration is highly inefficient and results in non-unique solutions. Here we propose a novel differentiable parameter learning (dPL) framework that efficiently learns a global mapping between inputs (and optionally responses) and parameters. Crucially, dPL exhibits beneficial scaling curves not previously demonstrated to geoscientists: as training data increases, dPL achieves better performance, more physical coherence, and better generalizability (across space and uncalibrated variables), all with orders-of-magnitude lower computational cost. We demonstrate examples that learned from soil moisture and streamflow, where dPL drastically outperformed existing evolutionary and regionalization methods, or required only ~12.5% of the training data to achieve similar performance. The generic scheme promotes the integration of deep learning and process-based models, without mandating reimplementation. Much effort is invested in calibrating model parameters for accurate outputs, but established methods can be inefficient and generic. By learning from big dataset, a new differentiable framework for model parameterization outperforms state-of-the-art methods, produce more physically-coherent results, using a fraction of the training data, computational power, and time. The method promotes a deep integration of machine learning with process-based geoscientific models.
Kanneganti et al merged UNOS and PHIS data to perform the first assessment of center‐level variation in acute rejection and biliary tract complications in the first year following pediatric liver transplantation (1). By combining these data sources, these authors have (1) created a novel, robust dataset for longitudinal study of pediatric liver transplant recipients and (2) have identified potentially new opportunities for quality improvement in post‐pediatric liver transplantation morbidity. In their analysis, Kanneganti et al found significant center‐level variation in acute rejection and, to a lesser degree, in biliary complications in the year following pediatric liver transplantation.
There is a lack of computing platforms to collect and analyze key data from traffic videos in an automatic and speedy way. Computer vision can be used in combination with parallel distributed systems to provide city authorities tools for automatic and fast processing of stored videos to determine the most significant driving patterns that cause traffic accidents while allowing to measure the traffic density. This study explores the integration of different tools such as parallel data processing, deep learning, and probabilistic models. We present an approach based on Convolutional Neural Network (CNN) and Kalman filters to detect and track vehicles captured by traffic cameras. To speed up analysis, we propose and evaluate a low-cost distributed infrastructure based on Hadoop and Spark frameworks and comprised of multicore CPU nodes for data processing. Finally, we present an algorithm to allow vehicle counting while avoiding inaccuracies generated when videos are split to be distributed for analysis. We found that it is possible to rapidly determine traffic densities, identify dangerous driving maneuvers, and detect accidents with high accuracy by using low-cost commodity cluster computing.
Water accounts for about 70% of the earth’s surface and is one of the most important sources of life. The degradation of the Salton Sea’s water quality due to excessive usage of fertilisers and pesticides in agriculture has posed serious risks to biodiversity, human health, and the local economy. As a result, it is necessary to establish effective and reliable methodologies for assessing water quality and predicting future trends. This work investigates regression and machine learning models such as linear regression, random forest, support vector machine (SVM) and long short-term memory (LSTM) to predict the salinity level of Salton Sea and forecast future trend. The proposed work studies the Salton Sea water quality and predict certain chemical or physical characteristics such as salinity, pH, and dissolved oxygen. It also finds the correlation between water indexes and environmental indexes surrounding the Salton Sea. The findings of this study can assist policymakers in determining the salinity of water and devising policies to minimise salt levels in order to manage freshwater for long-term development. While evaluating experiments, it has been found that the machine learning methodology is more flexible and accurate than traditional statistical methods, and that it aids in achieving better results.
Autonomous ships are promoted as the future of the maritime transport industry aiming to overcome conventional vessels in terms of performance, safety and environmental impact. Yet their tangled cyber-physical-social interactions and new emerging properties induce questions regarding their liability and trustworthiness. Digital simulations and sea trials are launched to assure the safety requirements and social expectations are met a priori. This paper presents the design of realistic testbed scenarios from huge historical data through a high-performance computational method to recommend a complete set of navigation scenarios for autonomy tests. The developed approach integrates traffic big data from Automatic Identification System (AIS) with high-resolution digital maps, vessel information registry, and digital nautical charts. All historical vessel-to-ground and vessel-to-vessel interactions are efficiently analyzed through a hierarchical method for collision and grounding conflicts assessment with a 15-minutes prediction horizon. Relative risk is evaluated accurately over full periods of predicted close-quarters situations subject to physical limits and sea-room availability for evasive maneuverers under COLREG rules and traffic separation schemes. Spatial dependencies among multiple conflicts define risky momentary traffic situations modelled through directed graph representation of nested interactions. Their temporal dependencies describe navigation scenarios through dynamic co-behaviors between multiple participating vessels over a period of time. Finally, we analyze negative/positive actions that increase/decrease the complexity. The presented algorithms are computationally very efficient, they scale to several (country*year)s where millions of scenarios are extracted, classified, and scored by their relative risk, complexity, and likelihood for firm post-test conclusions.
BACKGROUND There is currently a lack of data on stroke hospitalizations and long-term outcomes in China. Therefore, we investigated 12-month stroke fatality, disability, and recurrence rates after the first-ever stroke METHODS This was a prospective nationwide hospital-based cohort study. From August to September 2019, all patients with first-ever stroke (ischemic stroke [IS], intracerebral hemorrhage [ICH], and subarachnoid hemorrhage [SAH]) and with symptom onset within 14 days from 232 hospitals were included. Case fatality, disability, and recurrence rates for one year were estimated. RESULTS In total, 36250 first-ever stroke patients from 194 hospitals were recruited (median age was 65(IQR, 56-73) years and 61.4% were male). The rate of intravenous thrombolysis and endovascular treatment for IS were 9.5% and 4.4%, respectively. In-hospital death rate was 1.9% (95%CI: 1.7%-2.0%) for stroke inpatients, ranging from 0.9% (0.8%-1.1%) for IS to 5.1% (4.6%-5.6%) for ICH. The 12-month fatality rate was 8.6% (95%CI: 8.3%-8.9%) for discharged stroke patients, ranging from 6.0% (5.7%-6.3%) for IS to 17.7% (16.7%-18.7%) for ICH. The 12-month disability rate was 16.6% (95%CI: 16.2%-17.0%) for stroke survivors, ranging from 11.1% (9.3%-12.8%) for SAH to 29.2% (27.9%-30.4%) for ICH. The stroke recurrence rate was 5.7% (5.5%-6.0%) for stroke survivors, ranging from 2.5% (1.7%-3.3%) for SAH to 6.4% (6.0%-6.7%) for IS. CONCLUSION Our results support the hypothesis that the prognosis of Chinese stroke patients appears to have improved and is not very bad.