Predictive analysis using machine learning: Review of trends and methods

Abstract

Artificial Intelligence (AI) has been growing considerably over the last ten years. Machine Learning (ML) is probably the most popular branch of AI to date. Most systems that use ML methods use them to perform predictive analysis. This paper aims to conduct a literature review of trends and methods of machine learning used for predictive analysis. To do this, we carried out a collection of research papers from three scientific databases. We then considered selection criteria in order to study only papers published in the last five years, prioritizing those published in peer-reviewed scientific journals. This process led to the selection of 30 research papers that were considered for this review. The purpose of this study is to provide researchers, companies or anyone wishing to perform predictive analysis with clues that will enable them to choose the best ML method(s) according to its field of application, based on the latest research works in the literature. This study highlighted the most used methods by field of application: DT and ANN in education, LR, RF and DT in building, DT in botany, RF and ANN in social science and RF in medicine.

Full text

Predictive analysis using machine learning: Review
of trends and methods
1st Patrick Loola Bokonda
Department of Computer Science
Mohammed V University in Rabat. EMI, Siweb team
Rabat, Morocco
loola.bokonda@gmail.com
2nd Khadija Ouazzani-Touhami
Department of Computer Science
MINES-RABAT School
Rabat, Morocco
ouazzani@enim.ac.ma
3rd Nissrine Souissi
Department of Computer Science
MINES-RABAT School
Rabat, Morocco
souissi@enim.ac.ma
Abstract—Artificial Intelligence (AI) has been growing con-
siderably over the last ten years. Machine Learning (ML) is
probably the most popular branch of AI to date. Most systems
that use ML methods use them to perform predictive analysis.
This paper aims to conduct a literature review of trends and
methods of machine learning used for predictive analysis. To
do this, we carried out a collection of research papers from
three scientific databases. We then considered selection criteria
in order to study only papers published in the last five years,
prioritizing those published in peer-reviewed scientific journals.
This process led to the selection of 30 research papers that were
considered for this review. The purpose of this study is to provide
researchers, companies or anyone wishing to perform predictive
analysis with clues that will enable them to choose the best ML
method(s) according to its field of application, based on the latest
research works in the literature. This study highlighted the most
used methods by field of application: DT and ANN in education,
LR, RF and DT in building, DT in botany, RF and ANN in social
science and RF in medicine.
Keywords—artificial intelligence, machine learning methods,
predictive analysis, supervised learning, unsupervised learning,
semi-supervised learning, reinforcement learning, medicine
I. INTRODUCTION
For the past ten years Artificial Intelligence (AI) has ex-
perienced a renaissance and particularly Machine Learning
(ML) has been the subject of great attention. The ultimate
goal of AI is to make machines capable of performing tasks
previously considered intelligent [1], [2] better than humans.
Is this not the promise of a new revolution that will overturn
our relationship to work and knowledge [2] ?
While some see this as an unexpected opportunity to
improve our performance in several areas, others see it as
the danger of an adversary ready to take the place of man
in various sectors. But never mind, researchers, politicians,
companies and governments all use AI for a variety of reasons:
cancer control, voter profiling, search engines, programmatic
advertisements, soldiers’ weapons, weather, data processing ,
etc.
A common element in most AI use cases is prediction.
Prediction of high risk of cancer, profiling of voters most likely
to vote for a particular candidate, prediction of driver behavior
[3], prediction of videos and/or advertisements that may be
of interest to a particular person, etc. Predictive analysis is
increasingly used.
But a recurring question arises every time a predictive
analysis needs to be performed: Which method should be
used? This paper is a literature review of trends and methods
of ML used for predictive analysis in the most recent studies.
Since ML is a discipline with its own terminology, and in
order to familiarize researchers newly interested in ML with
concepts specific to this discipline, we have devoted section II
to the definition of terms used in ML. The aim is to facilitate
the understanding and use of the analysis and results of this
study.
The approach undertaken and explained in section III has
made it possible to select and study only the research carried
out over the last five years. The aim is to provide researchers,
companies or anyone wishing to perform predictive analysis
with clues to help them choose the best ML method(s) for
their field of application, based on the latest research in the
literature. Only papers from refereed journals were included
in this review.
The result of this study is presented in section IV. These
are the disciplines where ML methods are frequently used
for predictive analysis, the most used learning techniques, the
most used ML methods, and even the most used ML methods
by discipline. The result presented in this way does not dictate
which method to choose, but rather what other researchers
around the world are using in a specific discipline to do
predictive analysis.
Our prospects for future research and the conclusion of this
work are presented in Section V.
II. BACKGROUND
Machine Learning is defined by Melo Lima and Dursun
Delen in [4] as: ”A subset of artificial intelligence, which
is often applied when computing devices attempt to mimic
human cognitive functions related to learning and problem
solving processes in order to achieve ”optimal” results”. For
their part, Harleen Kaur and Vinita Kumari [5] define Machine
Learning as ”the development of algorithms and techniques
that enable computers to learn and acquire intelligence based
on past experience. This is a branch of Artificial Intelligence
(AI) and is closely related to statistics. Learning means that
the system is able to identify and understand the data entered,
so that it can make decisions and predictions based on them”.

In the context of this study, we use the following summary
definition: Machine Learning consists of training systems
capable of understanding the data entered in order to predict
responses or extract useful information from them. It is a sub-
set of artificial intelligence and is closely related to statistics.
In terms of the number of learning techniques that Machine
Learning understands, we have found that not all authors speak
with one voice on this subject. Some like Jorge Casta˜
n´
on
in [6] or Harleen Kaur and Vinita Kumari in [5] distinguish
between two types of learning, supervised and unsupervised.
Others such as Paul Lanier et al. in [7] consider three types of
learning, supervised, unsupervised and semi-supervised. Nirav
J. Patel and Rutvij H. Jhaveri in [8] consider reinforcement
learning to be the third type by removing the semi-supervised
from the list. Abdallah Moujahid et al. in [9] distinguish four
types of learning, supervised, unsupervised, reinforcement and
deep learning.
In this study, we consider four types of learning: supervised,
unsupervised, semi-supervised and reinforcement learning.
Supervised learning is used when historical data is avail-
able for a certain problem. The system is trained with the
respective inputs and responses, and then used to predict
responses for new inputs [5].
Supervised learning is subdivided into two sub-types: clas-
sification and regression [10].
Classification involves finding a relationship between dis-
crete inputs and discrete outputs. Output variables are also
called categories or labels. A mapping function (classifier)
is constructed by analyzing training data in the learning
step, and this classifier is adopted to predict categorical class
labels in the classification step [10]. Regression, on the other
hand, involves estimating or predicting continuous quantities.
Regression relies on input statistical characteristics to establish
the relationship between two or more independent variables,
[10].
Unsupervised learning, unlike supervised learning, does
not come with labels (no output vectors). The objective of
unsupervised learning is to analyze the structure of the data
and extract useful information from it without any explicit
indication of the expected result [10].
Unsupervised learning includes two sub-types: clustering
and dimensionality reduction [10].
clustering consists of dividing a set of objects into different
groups so that the objects in each group are as similar as
possible to each other, and the different groups are as differ-
ent as possible from each other [10]. While dimensionality
reduction aims to transform a large data space into a smaller
space without losing the useful information from the original
data [10].
Semi-supervised learning, as its name suggests, is a hybrid
of the two approaches mentioned above. Semi-supervised
learning is commonly used when some cases (problems) have
values for both covariates (inputs) and outcomes (outputs), but
the majority of cases have values only for the covariates and
lack data on the expected outcome [7].
Reinforcement learning is a particular area of Machine
Learning that is based on taking certain actions followed by
numerical rewards to achieve a goal. The important point
is that whoever undertakes an action, called an agent, in a
particular world, called an environment, does not know which
action is good or bad, but he will learn which ones will give
the greatest rewards by trying them out [9].
The word, technique will be used to refer to a type of learn-
ing (supervised, unsupervised, semi-supervised, or reinforce-
ment) [5], [11], [12]. For the term, method, it will be used to
refer to different methods of Machine Learning: ANN, SVM,
DT, etc. [13]. We will retain the nuance between algorithm
and model proposed by [13]. A model is a set of hypotheses
about a problem domain, expressed in a precise mathematical
form, which is used to create a Machine Learning solution
[13]. Whereas an algorithm is simply a set of instructions
used to implement a model to solve a problem or perform a
calculation.
III. STUDIED MACHINE LEARNING METHODS
This study focuses on the use of Machine Learning (ML)
methods in predictive analysis. To do this, we carried out a
research in three scientific databases: Science Direct, Springer
Link and IEEE Xplore. This search yielded 10 095 papers for
the three databases.
We then considered selection criteria in order to limit the
number of papers to only those published in the last five years,
prioritizing those published in peer-reviewed scientific journals
and using the relevance ranking tools provided by the scientific
databases.
At the end of the application of these criteria, 246 papers
stood out from the batch. Of these 246 papers, 13 duplicates
were detected and removed. The remaining 233 papers were
read in full and 30 were included in this study. This approach
allowed only the most recent publications to be retained. Of
the 30 papers, six were published in 2020, twenty-one in 2019
and three in 2018. Table I lists the papers considered in this
study, their years of publication and their fields of application.
We can see that ML methods have been used for predictive
analysis in several different areas. The papers studied cover
five domains: construction and botany for 3.3% each, educa-
tion for 10%, social science for 26.7% and medicine for 56.7%
of the papers.
Medicine being a very broad field, we have listed five
specialties in particular: cardiology [14], [15]; oncology [16],
[17], [18]; diabetology [11], [5], [19]; psychiatry [20] and
pediatric surgery [12], in addition to which it will be necessary
to add general medicine which groups together the largest
number of papers [21], [22], [14], [20], [23], [24], [25].
The review of these papers identified 23 ML methods
used to perform predictive analysis. We then organized these
methods by learning type, learning sub-type, and field of
application. Table II provides a summary of this organization.
IV. ANALYSIS AND DISCUSSION
As stated above, we distinguish four learning techniques.
It was observed during this study that supervised learning is

TABLE I
ST UDI ED PAP ER S BY YE AR AN D AP PLI CATI ON FI ELD
Paper Application field Year
Predicting adults likely to develop heart failure using readily available clinical information [15] Medicine - Cardiology
Using machine learning to predict opioid misuse among U.S. adolescents [26] Social Science
Machine learning models for credit analysis improvements: Predicting low-income families’ default
[27]
Social Science
A Mobile Application for Early Prediction of Student Performance Using Fuzzy Logic and Artificial
Neural Networks [28]
Education 2020
learning predictive model based on national data for fatal accidents of construction workers [29] Building
Testing the convergent- and predictive validity of a multi-dimensional belief-based scale for attitude
towards personal safety on public bus/ minibus for long-distance trips in Ghana: A SEM analysis [30]
Social Science
Predictors of length of stay in the coronary care unit in patient with acute coronary syndrome based
on data mining methods [14]
Medicine - Cardiology
Application of the Albumin-Bilirubin Grade in Predicting the Prognosis of Patients With Hepatocellular
Carcinoma: A Systematic Review and Meta-Analysis [16]
Medicine - Oncology
Computational models for predicting anticancer drug efficacy: A multi linear regression analysis based
on molecular, cellular and clinical data of oral squamous cell carcinoma cohort [17]
Medicine - Oncology
Machine-learning analysis of contrast-enhanced CT radiomics predicts recurrence of hepatocellular
carcinoma after resection: A multi-institutional study [18]
Medicine - Oncology
Predicting the botanical and geographical origin of honey with multivariate data analysis and machine
learning techniques: A review [31]
Botany
Can we predict lesion detection rates in second-look ultrasound of MRI detected breast lesions? A
systematic analysis [21]
General Medicine
Predictive analytics for hospital admissions from the emergency department using triage information
[22]
General Medicine
A predictive analytics framework for identifying patients at risk of developing multiple medical
complications caused by chronic diseases [32]
General Medicine
Using Machine Learning Applied to Real-World Healthcare Data for Predictive Analytics: An Applied
Example in Bariatric Surgery [11]
General Medicine
Predictive Analytics and Modeling Employing Machine Learning Technology: The Next Step in
Data Sharing, Analysis, and Individualized Counseling Explored With a Large, Prospective Prenatal
Hydronephrosis Database [12]
Medicine - Pediatric
Surgery
2019
Residential demand response program: Predictive analytics, virtual storage model and its optimization
[33]
Social Science
Predicting and explaining corruption across countries: A machine learning Approach [4] Social Science
Utilizing early engagement and machine learning to predict student outcomes [34] Education
Identifying predictors of probable posttraumatic stress disorder in children and adolescents with
earthquake exposure: A longitudinal study using a machine learning approach [35]
Social Science
Patient clustering improves efficiency of federated machine learning to predict mortality and hospital
stay time using distributed electronic medical records [23]
General Medicine
An automated machine learning-based model predicts postoperative mortality using readily-extractable
preoperative electronic health record data [24]
General Medicine
Educational data mining: Predictive analysis of academic performance of public-school students in the
capital of Brazil [36]
Education
Ensemble method based predictive model for analyzing disease datasets: a predictive analysis approach
[25]
General Medicine
A Predictive Analytics-Based Decision Support System for Drug Courts [37] Social Science
Preventing Infant Maltreatment with Predictive Analytics: Applying Ethical Principles to Evidence
Based Child Welfare Policy [7]
Social Science
Predictive models for diabetes mellitus using machine learning techniques [19] Medicine Diabetology
Predictive modelling and analytics for diabetes using a machine learning Approach [5] Medicine Diabetology
Processing electronic medical records to improve predictive analytics outcomes for hospital readmissions
[38]
General Medicine 2018
Applications of machine learning algorithms to predict therapeutic outcomes in depression: A meta-
analysis and systematic review [20]
Medicine - Psychiatry
the most widely used technique when it comes to predictive
analysis. We can see this in Fig. 1, which shows the rates of
use of the different learning techniques for the studied papers.
70% of the methods used to perform predictive analysis are
supervised learning, followed by unsupervised learning and
very little semi-supervised learning. None of the authors used
reinforcement learning to make prediction.
This observation gives a first clue in the choice of the type
of learning to do predictive analysis.
Another clue can be found in the fields of application of
these methods. Fig. 2 shows the areas of application of ML
methods for predictive analysis. We can see that medicine is
the most used discipline as an application area for ML methods
for predictive analysis. But it is not only medicine, there are
also social science, followed by construction, then botany and
finally education.
It should be noted that the use of a method in a field is
not exclusive. A method can be used in several areas. We see
for example the ANN which is used in medicine [14], [22],
[32], [5], [20], [23], in social science [26], [27], [4] and in
education [28]; or the CA which is used both in botany [31]
and in social science [33].

TABLE II
ML METHODS BY LEARNING TECHNIQUE,PAP ER AN D AP PLI CATI ON FIE LD
Method Description Type of learning Sub-Type Papers Application
fields
ANN Artificial Neural Network Supervised and Unsu-
pervised
Regression and cluster-
ing
[14] [22] [32] [26] [27]
[28] [4] [5] [20] [23]
Medicine,
Social Science,
Education
SVM Support Vector Machine Supervised Classification [14] [32] [4] [5] [20]
[22] [27] [25]
Medicine,
Social Science
DT Decision Tree Supervised Classification and Re-
gression
[14] [15] [31] [21] [32]
[27] [34] [29] [25] [19]
Medicine,
Building,
Education,
Botany
AC Auto Classifier Node Supervised Classification [14] Medicine
MLR Multi Linear Regression Supervised Regression [17] Medicine
Cox modele Supervised Regression [15] [18] Medicine
MRMR Maximum Relevance Minimum
Redundancy
Semi-Supervised [18] Medicine
RF Random Forest Supervised Classification and Re-
gression
[18] [22] [26] [4] [29]
[15] [24] [38] [25] [37]
[19]
Medicine,
Social Science,
Building
PCA Principal Component Analysis Unsupervised Dimentionnality Reduc-
tion
[31] [25] Medicine,
Botany
LDA Linear Discriminant Analysis Supervised Classification [31] [20] Medicine,
Botany
CA Cluster Analysis Unsupervised Clustering [31] [33] Social Science,
Bontany
XGBoost eXtreme Gradient Boosting al-
gorithm
Supervised Classification and Re-
gression
[22] [35] Medicine,
Social Science
LR Logistic Regression Supervised Regression [22] [32] [11] [29] [20]
[37] [19]
Medicine,
Building
GBM Gradient Boosting Machine Supervised Regression and classifi-
cation
[26] [20] [36] [19] Medicine
Fuzzy Supervised and Unsu-
pervised
Classification and Clus-
tering
[28] Education
RBF Radial Basis Function Supervised and Unsu-
pervised
Classification and Clus-
tering
[5] Medicine
K-NN K-Nearest Neighbour Supervised Classification [5] [25] Medicine
Adaboost Supervised Classification [29] Building
K-Means Clustering Unsupervised Clustering [23] Medicine
NB Na¨
ıve Bayes Supervised Classification [25] Medicine
PR Poisson Regression Supervised Regression [37] Social Science
OLSR Ordinary Least Squares Regres-
sion
Supervised Regression [37] Social Science
Bayesian meth-
ods
Supervised and Unsu-
pervised
Regression and Cluster-
ing
[7] Social Science
It is therefore important to have an overview of the most
commonly used ML methods for predictive analysis, all fields
combined. For this we have Fig. 3.
From Fig. 3, it is clear that RF is the most widely used ML
method for predictive analysis across all domains, followed
by ANN and DT, then SVM and then LR and the others.
That said, although RF is the most widely used method, it is
interesting to note its absence in some fields, such as education
and botany. The same is true for ANN which is not used in
building and botany, DT which is not used in social science
and botany, SVM which is not used in building, botany and
education, and LR which is not used in education and botany.
Thus, to make a good choice, one must consider the use
of methods in each area. For this reason, Fig. 4 shows the
usage rate of the five most commonly used ML methods for
predictive analysis by application domain. These five methods
alone cover 22 of the 30 papers, as well as the five application
areas identified in this study.
Fig. 4 shows, for the studied papers in this review, that:
•RF method is the most widely used in medicine;
•ANN and DT are the most used in education;
•DT is the only method used in the only botany paper
reviewed in this review;
•RF has the same usage rate as ANN in social science;
•LR, RF and DT have the same rate of use in building.
V. CONCLUSION AND FUTURE WORK
Following our previous research work [39], [40], [41], and
in order to select the most appropriate predictive analysis
methods to perform predictive analysis for epidemiological
diseases, we conducted a literature review.
This study presents a literature review of trends and methods
of ML used for predictive analysis. In order to provide a
review that reflects the current state of research, we adopted

Fig. 1. ML techniques used for predictive analysis in studied papers
Fig. 2. Fields of application of ML methods for predictive analysis
an approach that allowed us to select only the most recent
research papers in the literature.
This made it possible to identify: The most used disciplines
as application fields of ML methods for predictive analysis:
medicine, social science, building, botany and education; The
most commonly used learning techniques: Supervised and
Unsupervised; The most commonly used ML methods: RF,
ANN, DT, SVM and LR; And the most used ML methods by
discipline.
The result presented in this way does not dictate which
method to choose, but rather what methods other researchers
around the world are using in a specific discipline to do pre-
dictive analysis. This can serve as an indicator for researchers
interested in predictive analysis. It should be note that the use
of ML method could depend closely on the specification of
Fig. 3. Rate of use of ML methods in studied papers
Fig. 4. Rate of use of the main ML methods by application fields
the problem regardless of its domain.
For our part, we plan to integrate ML methods into a mobile
data collection system to perform predictive analysis for an
epidemiological investigation. For example an investigation
about covid-19.
REFERENCES
[1] Montreal University:
http://www.iro.umontreal.ca/ nie/IFT3335/Introduction.html, accessed
on 13 june 2020.
[2] CNRS LE JOURNAL, Yaroslav Pigenet :
https://lejournal.cnrs.fr/articles/des-machines-enfin-intelligentes.
[3] Zinebi, K., Souissi, N., and Tikito, K. (2018). Driver Behavior Analysis
Methods: Applications oriented study. In Proceedings of the 3rd Inter-
national Conference on Big Data, Cloud and Application (BDCA 2018),
April 2018 in Kenitra Morocco.

[4] Lima, M. S. M., and Delen, D. (2020). Predicting and explaining
corruption across countries: A machine learning approach. Government
Information Quarterly, 37(1), 101407.
[5] Kaur, H., and Kumari, V. (2018). Predictive modelling and analytics
for diabetes using a machine learning approach. Applied computing and
informatics.
[6] Casta˜
n´
on, J. (10). Machine Learning Methods that Every Data Scientist
Should Know. Consultado em Outubro, 16, 2019.
[7] Lanier, P., Rodriguez, M., Verbiest, S., Bryant, K., Guan, T., and Zolotor,
A. (2020). Preventing infant maltreatment with predictive analytics: ap-
plying ethical principles to evidence-based child welfare policy. Journal
of family violence, 35(1), 1-13.
[8] Patel, N. J., and Jhaveri, R. H. (2015, January). Detecting packet
dropping nodes using machine learning techniques in Mobile ad-hoc
network: A survey. In 2015 International Conference on Signal Process-
ing and Communication Engineering Systems (pp. 468-472). IEEE.
[9] Moujahid, A., Tantaoui, M. E., Hina, M. D., Soukane, A., Ortalda, A.,
ElKhadimi, A., and Ramdane-Cherif, A. (2018, June). Machine learning
techniques in ADAS: A review. In 2018 International Conference on
Advances in Computing and Communication Engineering (ICACCE)
(pp. 235-242). IEEE.
[10] Yang, H., Xie, X., and Kadoch, M. (2020). Machine Learning Tech-
niques and A Case Study for Intelligent Wireless Networks. IEEE
Network, 34(3), 208-215.
[11] Johnston, S. S., Morton, J. M., Kalsekar, I., Ammann, E. M., Hsiao, C.
W., and Reps, J. (2019). Using machine learning applied to real-world
healthcare data for predictive analytics: an applied example in bariatric
surgery. Value in Health, 22(5), 580-586.
[12] Lorenzo, A. J., Rickard, M., Braga, L. H., Guo, Y., and Oliveria,
J. P. (2019). Predictive Analytics and Modeling Employing Machine
Learning Technology: The Next Step in Data Sharing, Analysis, and
Individualized Counseling Explored With a Large, Prospective Prenatal
Hydronephrosis Database. Urology, 123, 204-209.
[13] Model-based Machine Learning:
http://www.mbmlbook.com/Introduction.html
[14] Rezaianzadeh, A., Dastoorpoor, M., Sanaei, M., Salehnasab, C., Moham-
madi, M. J., and Mousavizadeh, A. (2020). Predictors of length of stay
in the coronary care unit in patient with acute coronary syndrome based
on data mining methods. Clinical Epidemiology and Global Health, 8(2),
383-388.
[15] Bergsten, T. M., Nicholson, A., Donnino, R., Wang, B., Fang, Y., and
Natarajan, S. (2020). Predicting adults likely to develop heart failure
using readily available clinical information: An analysis of heart failure
incidence using the NHEFS. Preventive Medicine, 130, 105878.
[16] Xu, L., Wu, J., Lu, W., Yang, C., and Liu, H. (2019, December).
Application of the Albumin-Bilirubin Grade in Predicting the Prognosis
of Patients With Hepatocellular Carcinoma: A Systematic Review and
Meta-Analysis. In Transplantation Proceedings (Vol. 51, No. 10, pp.
3338-3346). Elsevier.
[17] Robert, B. M., Brindha, G. R., Santhi, B., Kanimozhi, G., and Prasad, N.
R. (2019). Computational models for predicting anticancer drug efficacy:
A multi linear regression analysis based on molecular, cellular and
clinical data of oral squamous cell carcinoma cohort. Computer methods
and programs in biomedicine, 178, 105-112.
[18] Ji, G. W., Zhu, F. P., Xu, Q., Wang, K., Wu, M. Y., Tang, W. W.,
... and Wang, X. H. (2019). Machine-learning analysis of contrast-
enhanced CT radiomics predicts recurrence of hepatocellular carcinoma
after resection: A multi-institutional study. EBioMedicine, 50, 156-165.
[19] Lai, H., Huang, H., Keshavjee, K., Guergachi, A., and Gao, X. (2019).
Predictive models for diabetes mellitus using machine learning tech-
niques. BMC endocrine disorders, 19(1), 1-9.
[20] Lee, Y., Ragguett, R. M., Mansur, R. B., Boutilier, J. J., Rosenblat,
J. D., Trevizol, A., and Chan, T. C. (2018). Applications of machine
learning algorithms to predict therapeutic outcomes in depression: a
meta-analysis and systematic review. Journal of affective disorders, 241,
519-532.
[21] Bumberger, A., Clauser, P., Kolta, M., Kapetas, P., Bernathova, M.,
Helbich, T. H., ... and Baltzer, P. A. (2019). Can we predict lesion de-
tection rates in second-look ultrasound of MRI-detected breast lesions?
A systematic analysis. European journal of radiology, 113, 96-100.
[22] Araz, O. M., Olson, D., and Ramirez-Nafarrate, A. (2019). Predictive
analytics for hospital admissions from the emergency department using
triage information. International Journal of Production Economics, 208,
199-207.
[23] Huang, L., Shea, A. L., Qian, H., Masurkar, A., Deng, H., and Liu,
D. (2019). Patient clustering improves efficiency of federated machine
learning to predict mortality and hospital stay time using distributed elec-
tronic medical records. Journal of biomedical informatics, 99, 103291.
[24] Hill, B. L., Brown, R., Gabel, E., Rakocz, N., Lee, C., Cannesson,
M., and Maoz, U. (2019). An automated machine learning-based model
predicts postoperative mortality using readily-extractable preoperative
electronic health record data. British Journal of Anaesthesia, 123(6),
877-886.
[25] Ramesh, D., and Katheria, Y. S. (2019). Ensemble method based
predictive model for analyzing disease datasets: a predictive analysis
approach. Health and Technology, 9(4), 533-545.
[26] Han, D. H., Lee, S., and Seo, D. C. (2020). Using machine learning to
predict opioid misuse among US adolescents. Preventive medicine, 130,
105886.
[27] de Castro Vieira, J. R., Barboza, F., Sobreiro, V. A., and Kimura,
H. (2019). Machine learning models for credit analysis improvements:
Predicting low-income families’ default. Applied Soft Computing, 83,
105640.
[28] Nosseir, A., and Fathy, Y. (2020). A Mobile Application for Early
Prediction of Student Performance Using Fuzzy Logic and Artificial
Neural Networks.
[29] Choi, J., Gu, B., Chin, S., and Lee, J. S. (2020). Machine learning pre-
dictive model based on national data for fatal accidents of construction
workers. Automation in Construction, 110, 102974.
[30] Sam, E. F., Brijs, K., Daniels, S., Brijs, T., and Wets, G. (2020). Testing
the convergent-and predictive validity of a multi-dimensional belief-
based scale for attitude towards personal safety on public bus/minibus
for long-distance trips in Ghana: A SEM analysis. Transport policy, 85,
67-79.
[31] Maione, C., Barbosa Jr, F., and Barbosa, R. M. (2019). Predicting
the botanical and geographical origin of honey with multivariate data
analysis and machine learning techniques: A review. Computers and
Electronics in Agriculture, 157, 436-446.
[32] Talaei-Khoei, A., Tavana, M., and Wilson, J. M. (2019). A predictive an-
alytics framework for identifying patients at risk of developing multiple
medical complications caused by chronic diseases. Artificial Intelligence
in Medicine, 101, 101750.
[33] Basnet, S. M., Aburub, H., and Jewell, W. (2019). Residential demand
response program: Predictive analytics, virtual storage model and its
optimization. Journal of Energy Storage, 23, 183-194.
[34] Gray, C. C., and Perkins, D. (2019). Utilizing early engagement and
machine learning to predict student outcomes. Computers & Education,
131, 22-32.
[35] Ge, F., Li, Y., Yuan, M., Zhang, J., and Zhang, W. (2020). Identifying
predictors of probable posttraumatic stress disorder in children and
adolescents with earthquake exposure: a longitudinal study using a
machine learning approach. Journal of affective disorders, 264, 483-493.
[36] Fernandes, E., Holanda, M., Victorino, M., Borges, V., Carvalho, R., and
Van Erven, G. (2019). Educational data mining: Predictive analysis of
academic performance of public-school students in the capital of Brazil.
Journal of Business Research, 94, 335-343.
[37] Zolbanin, H. M., Delen, D., Crosby, D., and Wright, D. (2019). A
Predictive Analytics-Based Decision Support System for Drug Courts.
Information Systems Frontiers, 1-20.
[38] Zolbanin, H. M., and Delen, D. (2018). Processing electronic medical
records to improve predictive analytics outcomes for hospital readmis-
sions. Decision Support Systems, 112, 98-110.
[39] Loola Bokonda P.,Ouazzani-Touhami K., and Souissi N. (2020) Mobile
Data Collection Using Open Data Kit. In: Serrhini M., Silva C., Aljahdali
S. (eds) Innovation in Information Systems and Technologies to Support
Learning Research. EMENA-ISTL 2019. Learning and Analytics in
Intelligent Systems, vol 7. Springer, Cham.
[40] Bokonda, P. L., Ouazzani-Touhami, K., and Souissi, N. (2019). Open
Data Kit: Mobile Data Collection Framework For Developing Countries.
8. 4749-4754. 10.35940/ijitee.L3583.1081219.
[41] Bokonda, P. L., Ouazzani-Touhami, K., and Souissi, N. (2020). A
Practical Analysis of Mobile Data Collection Apps. International Journal
of Interactive Mobile Technologies (iJIM), Vol. 14, No. 13, 2020. In
press