Machine Learning for Personal Credit Evaluation: A Systematic
Review
CANO CHUQUI JORGE
Faculty of Engineering and Architecture
Universidad Privada César Vallejo
Av. Del Parque 640, San Juan de Lurigancho 15434
PERÚ
OGOSI AUQUI JOSÉ ANTONIO
Faculty of Engineering
Universidad Tecnológica del Perú
Av. Arequipa 265, Cercado de Lima 15046
PERÚ
Abstract: - The importance of information in today's world as it is a key asset for business growth and
innovation. The problem that arises is the lack of understanding of knowledge quality properties, which leads to
the development of inefficient knowledge-intensive systems. But knowledge cannot be shared effectively
without effective knowledge-intensive systems. Given this situation, the authors must analyze the benefits and
believe that machine learning can benefit knowledge management and that machine learning algorithms can
further improve knowledge-intensive systems. It also shows that machine learning is very helpful from a
practical point of view. Machine learning not only improves knowledge-intensive systems but has powerful
theoretical and practical implementations that can open up new areas of research. The objective set out is the
comprehensive and systematic literature review of research published between 2018 and 2022, these studies
were extracted from several critically important academic sources, with a total of 73 short articles selected. The
findings also open up possible research areas for machine learning in knowledge management to generate a
competitive advantage in financial institutions.
Key-Words: machine learning, credit scoring, risk assessment, algorithms, artificial intelligence.
Received: March 25, 2021. Revised: April 14, 2022. Accepted: May 12, 2022. Published: July 1, 2022.
1 Introduction
Time analysis is critical because financial
institutions consistently implement the credit
scoring model over time. Therefore, due to the
complexity and flexibility of the training process,
ML methods can be more sensitive to overtime
disturbances. It is susceptible to overfitting
problems and may become unstable over time.
In addition, there is a lack of ability to account
for time complexity in actual business operations.
Everything depends on the model life cycle, from
data collection, model development, and validation
to the final model. This model can present new
challenges and uncertainties to the generalization of
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
62
Volume 10, 2022
GUADALUPE MORI, VICTOR HUGO
Faculty of Engineering
Universidad Privada San Juan Bautista
Ex Hacienda Villa, Av. José Antonio Lavalle s/n, Chorrillos 15067
PERÚ
OBANDO PACHECO, DAVID HUGO
Faculty of Engineering
Universidad Peruana de Ciencias Aplicadas
Prolongación Primavera 2390, Lima 15023
PERÚ
ML models over time. ML production
representations are generally scheduled for the
model to be applied at least one year after the
development data period.
The motivation for this study stems from the fact
that to access loans, applicants must remain enrolled
in a regular socioeconomic dimension essential to
repay the money. Applicants are concerned about
whether they are good repayers and to ensure that
they can meet the requirements they expect based on
their client activities. The study contributes not only
to the collection of citations and empirical
evaluation of studies on personal credit evaluation
but also addresses the measurement of Machine
Learning methods to influence bank lending
processes in a formalized financial institution with
cash flow, which provides a particular interest rate
and certain loan application requirements.
This survey includes empirical reviews of
personal credit score surveys and extends the
literature that incorporates information provided by
banks or financial institutions regarding the services
offered. This article is divided into Chapter 1-
Introduction, Chapter 2-Methodology, Chapter 3-
Results a, and Chapter 4-Conclusions for future
work.
2 Methodology
This paper conducts a research study on the personal
credit scoring process using machine learning. This
work is called a preliminary survey because it is the
first time it has been conducted. We believe that this
review is part of an evidence-based personal credit
assessment. It aims to collect research that can be
used to implement machine learning methods,
models, or approaches. The review articles are
elaborated based on a methodology, which gives us
very clear steps to make a review article, that
methodology is called the "Systematic Literature
Review", it is a methodology to make articles, and
therefore, it is a starting point like all research, to
establish the problems of something unknown or
something that you want to optimize, and the
objectives that you want to achieve. The review tells
us what are the problems we have and what things
we want to achieve as point number one, then, we
look for good sources of information, because it is a
systematic review of the literature, but in this case,
the literature that we review, are scientific articles,
we do not review web pages, books, or theses for
this work, as point number two. Then, we have to do
a systematic, orderly, and rigorous search with
search equations. We can find hundreds, thousands,
or hundreds of thousands, but it has to be done with
a systematic, orderly and rigorous filter, and we will
be left with 70 to 80 articles as point number three.
After these articles are reviewed, we answer the
questions as point number four. When we generate
the answers to these questions, we will have the
fundamental input to prepare the review article,
which has all the parts of an article as point number
five; its title, its authors, its affiliation, its summary,
its keywords, the sources of information, likewise, it
has the result of showing the answers raised in point
number one. This is the overview of how review
articles are developed, as shown in Figure 1.
Fig. 1: SRL methodology.
2.1 Formulation of Questions and Objectives
There are two specific problems tracked in the meta-
analysis presented in Table 1.
Table 1. Research questions
Research Question
Motivation
Motivation
RQ1: What are the methods
used to apply machine
learning for personal credit
assessment?
Demonstrate which methods
are used to apply machine
learning for personal credit
evaluation.
RQ2: What complementary
methods are used to apply
machine learning for personal
credit assessment?
Demonstrate which
complementary methods
apply machine learning for
personal credit assessment.
2.2 Search Equation
2.2.1 Search Sources
Systematic literature searches take information from
11 search sources and use search terms to create
general research equations.
Taylor & Francis
ProQuest
WOS
IEEE Xplore
ScienceDirect
Scopus
EBSCOhost
Wiley Online Library
IOP
ERIC
Google Scholar
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
63
Volume 10, 2022
2.2.2 Search Equations
The search terms used in this study, applying the
terms of use of each source, are shown in Table 2.
Table 2. Search Equations
Fuente
Taylor &
Francis
ProQuest
WOS
IEEE Xplore
ScienceDirect
Scopus
EBSCOhost
Wiley Online
Library
IOP
ERIC
Google
Scholar
2.3 Studies Identified
2.3.1 Integrated Chart of Numbers of Results by
Source
As shown in Figure 2, "n" was found in the 11-
source search formula where (N = 164,793) articles
were published in 11 sources. The overall results
published in 11 sources representing the total
number of pieces sampled were reported.
Fig. 2: Consolidated Graph of the Number of
Results by Source.
2.3.2 Consolidated Matrix of Number of Results
by Source
Table 3 shows two search terms for each source, and
you can apply the search terms and their logical
relationships to get the total for each reference. This
will eventually show 11 stars with a total result of
164,793 for the articles found.
Table 3. Consolidated matrix of the resulting
number by source
Source
Generic research equation
Total
Taylor &
Francis
[All: machine learning] AND [All:
personal credit evaluation] AND
[[All: method] OR [All:
methodology] OR [All: model]]
13,556
ProQuest
(Machine Learning) AND
(Personal Credit Evaluation) AND
((method OR methodology OR
model))
135,346
WOS
(Machine Learning) AND
(Personal Credit Evaluation) AND
(method OR methodology OR
model)
9
IEEE Xplore
("All Metadata":Machine Learning)
AND ("All Metadata":Personal
Credit Evaluation) AND ("All
Metadata":method OR "All
Metadata":methodology OR "All
Metadata":model)
215
ScienceDirect
("Machine Learning") AND
("Personal Credit Evaluation")
AND ("method" OR
"methodology" OR "model")
9
Scopus
ALL ( ( machine AND learning )
AND ( personal AND credit
AND evaluation ) AND ( (
method OR methodology OR
model ) ) )
2,225
EBSCOhost
TX Machine Learning AND TX
Personal Credit Evaluation AND
TX ( method OR methodology OR
model )
64
Wiley Online
Library
"Machine Learning" anywhere and
"Personal Credit Evaluation"
anywhere and "method OR
13,208
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
64
Volume 10, 2022
methodology OR model" anywhere
IOP
Machine Learning AND Personal
Credit Evaluation AND (method
OR methodology OR model)
6
ERIC
Machine Learning AND (Personal
Credit Evaluation) AND (method
OR methodology OR model)
42
Google
Scholar
("Machine Learning" AND
"Personal Credit Evaluation" AND
method OR methodology OR
model)
219
2.4 Exclusion Criteria
2.4.1 Exclusion Criteria
To avoid selection bias and strictly apply the
criteria, we created a list of objective exclusion
criteria. As we apply these criteria, we are leaving
out a large number of articles. When we finish the
last criterion, we are left with a more manageable
number of documents, to have around 80 articles
that will serve to answer the questions of this
research with distance and objectivity. These criteria
depended on the study objectives, for which a total
of 8 exclusion criteria were identified, as shown in
Table 4 below.
Table 4. Exclusion Criteria
N
Descripción
Does it
comply?
CE1
Articles are older than five years O
CE2
Articles are not written in the English
language OR
CE3
The documents are not of article type.
CE4
The titles and keywords of the articles are
not very appropriate.
CE5
The articles are not unique.
CE6
The abstract of the articles is not very
relevant O
CE7
The reports do not mention a
methodology, model, or method O
CE8
The proposed solution does not apply to
Customer Service OR
Teaching_Learning, Recruitment, OR
Political campaign management.
X
Of the eight exclusion criteria, five were selected to
evaluate and investigate the articles found.
When applying the criteria:
CE1: excluded 897 items older than five
years, resulting in 51,896 items using the
first criterion.
In CE2, we excluded 1,903 items that were
not OR in English and used a second
criterion to exclude 49,993 items.
CE3: 38,238 items were excluded instead of
article-type documents for 11,755 items
when the third criterion was applied.
CE4: 11,664 items were excluded, their
titles and keywords were not very
appropriate for the article, and 91 items
were obtained using the fourth criterion.
CE5: 18 items are excluded and ambiguous;
applying the fifth criterion totals 73 items
and processing them for research.
2.4.2 The Aggregate Number of Results when
Exclusion Criteria Apply
Table 5 shows the primary studies. The filters
applied have two exclusions for each (Filter 1:
CE1CE2, Filter 2: CE3CE4, and Filter 3: CE5), with
the final filter being the element after maintaining
the exclusion criteria.
Table 5. The aggregate number of results when
exclusion criteria apply
Source
Initials
Filter
1:
CE1-
CE2
Filter
2: CE3-
CE4
Filter
3: CE5
Taylor & Francis
13,556
1,732
1,545
6
ProQuest
135,346
46,459
8.146
16
WOS
9
9
8
4
IEEE Xplore
215
8
8
8
ScienceDirect
9
7
7
2
Scopus
2,225
1,204
702
25
EBSCOhost
64
18
18
7
Wiley Online
Library
13,208
2,060
1,313
3
IOP
6
6
4
2
ERIC
42
3
2
0
Google Scholar
219
219
2
0
Total
164,793
51725
11753
73
2.5 Review of the Studies
2.5.1 The Aggregate Number of Results when
Exclusion Criteria Apply
In this section, the answers that appear in each study
are identified from the full-text studies so that the
page on which the answers appear can be viewed
promptly, where the conceptual clarification of the
research topic can also be started, such as it is
mentioned in Table 6.
Table 6. The aggregate number of results when
exclusion criteria apply
N
RQ1
RQ2
RQ3
RQ4
RQ5
RQ6
[1]
6,12,
14
1
1,2
1
10,14,1,8
11,16
[2]
1,4
1
1
1
2
[3]
1
1
2
[4]
1
1
1
3
[5]
1
[6]
2,11
1
1
1
3
2,3,8
[7]
1
1
2
[8]
1
1
1
2
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
65
Volume 10, 2022
[9]
3
1
2
1
6
8
[10]
1
1
1
[11]
1
1
1
6
[12]
1
[13]
1
1,2
1
13,14
4
[14]
1
2
[15]
1
1
1
[16]
1
1
2
1,2
[17]
8
1
1
1
[18]
1
1
1
[19]
1
1
1
6
[20]
5
1
1
1
1
3
[21]
1
1
1
[22]
2,3,4
1
1,2
2
6
1
[23]
1
1
1
1
[24]
1
1
1
1
3
[25]
1
1
2,3
[26]
1
1
1
2
[27]
1
1,2
1
4
4
[30]
1
1
1
6
3
[31]
1
1
1
5
[32]
2
2
2
[33]
1
1
1
[34]
10
1
1
1
6
12
[35]
1
1
1
[36]
1,2
1
1
1
1,2
[37]
1
[38]
2
1
1,2
1
2,3,4
[39]
1
1
1
[40]
1
1
1
[41]
2
1
1
1
[42]
1
1
1
2
[43]
6
1
1,2
1
9
[44]
1
1
1
2,3
[45]
1
1
1
9
[46]
1,2
1
1
1
2,3
[47]
2,3
1
1
1
4,5
[48]
1
1
1
[49]
1
1
1
5
4
[50]
1
1,2
1
1,4,6,7
[51]
8,12
1
1,2
1
6,8,10,11,
[52]
3,7,1,0
12,14
[53]
1
1,2
1
5,7,11
7,10
[54]
16
1
2
4,16
[55]
4
1
1,2
1
6
4
[56]
1
1,2
1
6
14,19
[57]
1
1,2
1
16
17
[58]
8,9
1
1
1
1,2
[59]
1
2
1
1
9,12
8
[60]
1
1
1
[61]
3,4
1
1,2
[63]
1
1
1
[64]
2
1
1
1
14,15
[65]
1
1
1
[66]
2,3
1
1
1
2
[67]
4
1
1,2
1
4,8,9,10
[68]
1
1
1
2
[69]
1
1
1
1,2
[70]
1
1
1,2
4
[71]
1
1
1
4
[72]
1
1,2
1
5,6,7
[73]
2
1
2
3 Results
Answers to RQs.
RQ1: What are the methods used to apply
machine learning for personal credit assessment?
Table 7. Definitions of most commonly used ML
methods for credit evaluation
N
ML Method
Definitions
References
Total
number of
articles
1
Supervised
learning
[4] [5] [7] [16] [19] [33]
[34] [37] [40] [41] [45]
[46] [49] [53] [63] [68]
16 (22%)
2
Unsupervised
learning
[1] [5] [7] [10] [13] [16]
[27] [28] [31] [37] [39]
[41] [46] [56] [61] [63]
[66] [69] [70] [78]
20 (27%)
The definition of supervised learning is used in
20 (27%) items and also emphasizes complementary
methods. Finally, the few articles mentioned it,
totaling 16 (22%). The following are the most
commonly used machine learning (ML) methods.
The meanings used to define ML methods are
unsupervised learning and unsupervised learning.
RQ2: What complementary methods are used to
apply machine learning for personal credit
assessment?
Due to the advanced technology associated with
big data, data availability, and computational power,
most banks or credit unions are innovating their
business models [1]. Credit scoring analytics is an
effective technique for assessing credit risk and is
one of the major research areas in the banking
industry [2]. Neural networks are one of the most
widely used methods to obtain credit scores [3].
Using machine learning through modeling and
prediction, both models were trained on real credit
card transaction datasets [4]. Currently, these
functions are performed manually and are subject to
expert evaluation [5]. Available multiclass
classifiers, such as random forest algorithms, can
perform this task very well, using available
customer data [6]. He suggested using this credit
scoring from unconventional data sources for online
lenders to enable them to investigate and detect
changes in customer behavior over time and target
unsecured customers based on their claims data [7].
In a rapidly developing economy, credit plays a very
important role. Several prediction models have been
developed to predict credit risk with many different
variables [8]. The information provided by the
candidates constitutes the variables of our analysis
[9]. Machine learning algorithms have dominated
many different industries [10]. Several aspects need
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
66
Volume 10, 2022
to be taken into account to make credit scoring
models understandable and to provide a framework
for making "black box" machine learning models
transparent, audible, and solvable [11]. With the
rapid development of corporate lending in China,
the creation of a high-risk credit scoring system has
become an important measure of financial
guarantees [12]. Non-performing loans are a serious
problem in the banking sector. Credit rating models
used logistic regression and linear discriminant
analysis to identify potential defaulters [13].
Financial institutions are faced with the need to
assess the creditworthiness of the borrower applying
for the loan. The best results are observed for
randomization [14]. The best prediction results are
obtained using conventional synthesis techniques,
namely, packing, random forest, and Bo
improvement [15]. Replacing subjective analysis
with objective credit analysis using deterministic
models would benefit Brazilian credit unions [16].
The model quadruples the accepted default rate to
break even from 8% to 32% [17]. Several 'rejection
inference' methods attempt to exploit the available
data for candidates that were rejected during the
learning process [18]. It then quantifies each
indicator and defines criteria for evaluating the
assessment results [19]. It then quantifies each
indicator and defines criteria for evaluating the
assessment results [20]. A good credit rating
decision support system allows telecom operators to
measure the creditworthiness of subscribers in detail
[21]. Current research on credit scoring in
microfinance is limited to genetic and regression
algorithms, which excludes newer machine learning
algorithms [22]. General psychometric modeling is
effective in predicting lifetime consumer mortgage
behavior [23]. The development of accurate
analytical credit scoring models has become an
important goal of financial institutions [24].
Choosing the optimal techniques, whether attribute
selection techniques, attribute assignment
techniques, or ML resampling mechanisms and
classifiers to support the coverage decision is
challenging and doable. The integrated SVM-
Logistic model is complementary and has a high
evaluation density [26]. For domain adaptation
problems, transfer learning techniques are often
used; however, it is very difficult to run accurate
predictions of unknown domain datasets in CSM
because name distributions may be different
depending on domain properties [27]. The
Dempster-Shafer synthesis method allows accurate
labeling by exploiting the advantages of both
methods [28]. Rural credit is one of the most
important inputs for agricultural production in the
world. However, it is the banking or non-banking
institutions that will decide how to apply this
advanced technology to reduce human biases in the
credit decision-making process [29]. This particular
model performs better than multilayer
backpropagation networks, probabilistic neural
networks, radial basis functions, and regression
trees, as well as other advanced classifiers [30]. A
credit default prediction model based on a complex
graph network can reflect nonlinear relationships
between borrower characteristics and default risk
and higher-order relationships between borrowers
[31]. The XGB model has obvious advantages in
both feature selection and classification
performance over logistic regression and the other
three tree-based models [32]. Application of radial-
based neural network model along with optimal
segmentation algorithm in credit scoring model of
personal loans to banks or other financial
institutions [33]. In a nonlinear method that
eliminates the obvious subjective and artificial
factors, the evaluation results are more objective and
effective [34]. The multiple averaging methods can
effectively reduce the diversity of the results, and
the accuracy will not be significantly reduced by the
different proportions of training and prediction sets
[35]. In emerging markets, there is a gap between
having a credit rating or credit score and having no
credit history [36]. It assists borrowers in the
fundraising process, allowing any number and size
of lenders to participate [37]. The training of the
model will be performed using machine learning-
based algorithms such as; Random Forest, Extreme
Gradient, Boost, Mild Gradient Boost, Adaboost,
and ExtraTrees [38]. Measuring credit risk is
essential for financial institutions due to the high
level of risk associated with bad credit decisions.
The recent Basel Accord specifies that reserve
requirements have increased according to risk [39].
A credit score is a central component of a
corporation's lending. The combination of credit
scoring and machine learning can integrate a
relatively complete functionality into the credit
scoring process [40]. The model structure is
determined by hyperparameters, aiming to address
the time-consuming and labor-intensive manual
adjustment problem, and the optimization method is
used for adjustment [41]. The SCSRF model
parameters were optimized by grid search [42].
Credit ratings are becoming increasingly important
in the financial sector. By changing the number of
nodes in a Spark cluster, the execution time of these
algorithms is compared and the analysis of variance
compares the execution time of each algorithm with
an increasing number of nodes [43]. To reduce the
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
67
Volume 10, 2022
negative impact of the unbalanced dataset on the
performance of the credit rating model, the SMOTE
technique was used to rebalance the target training
dataset [44]. There are relatively few transparency
models that take interpretability and clarity into
account [45]. To identify eligible end customers
from defaulters, a credit scoring model is used to
reliably screen the credit data using a combination
of Min-Max normalization and linear regression.
[Forty-six]. The support vector machine is the most
widely used classifier for credit rating and, although
the system performs well, it does not apply
collateral approaches [47]. Corporate insolvency has
significant negative effects on the economy. The RF
algorithm shows utility in credit risk management
[49]. Maximum machine learning is used as a
scoring tool for the credit risk assessment model
[50]. Intensive machine learning enables multilayer
neural networks to perform operations to facilitate
operations and business dynamics [51]. The
proposed P2P personal credit scoring model is
superior to both individual models and other sets of
criteria [52]. The influence of controlling
shareholder characteristics on corporate risk has
been a popular topic of debate in academic and
theoretical circles [53]. Online personal loans are a
new form of lending [54]. The method assigns terms
to the embedding space, groups the linguistically
related terms into semantic clusters, and then selects
the flexible semantic items corresponding to the
semantic clusters [55]. Object selection techniques
or object selection techniques should be
incorporated in predictive model building to
improve prediction performance [56]. According to
the application scenario of credit scoring of personal
credit data, the test data set is cleaned, the separated
data is coded as HOT and the data is normalized
[57]. The incorporation of macroeconomic variables
can improve the performance of existing models
[58]. To confirm the effectiveness of the proposed
credit rating model, experiments were conducted
with realistic credit data sets for comparison [59].
The accuracy and kappa values for all four methods
exceeded 90 %, and RF outperformed other rating
models [60]. Three hybrid AI models have been
studied, including decision tree: artificial neural
network, decision tree: logistic regression, and
decision tree: dynamic Bayes network [61]. Online
financial institutions lack effective methods to
evaluate personal credit, which seriously hinders the
development of personal credit business [62]. More
and more attention is paid to the use of machine
learning algorithms to predict people's credit ratings
in the era of artificial intelligence [63]. Because
there are many unusual users in these data, they are
"real but fake data" of personal credit rating [64].
The model built there performed well on some of
the scales used to compare it to other commonly
used raters [65]. Blockchain, decision trees, and
other technologies can effectively improve the
transparency of personal credit information in the
field of Internet finance [66]. Based on the seed
neural network, a predictive model of the
probability of granting formal credit to farmers was
built [67]. Geospatial data collection from location-
based services can provide location evidence during
spatial information analysis [68]. Due to the rapid
increase in the number of personal loan applications,
the importance of credit risk assessment for
practitioners and researchers is increasing [69].
Using blockchain, decision trees, and other
technologies, this paper designs the credit rating
process and establishes the individual credit rating
technology [70]. CNN is used to create a model to
predict individual credit default, and ACC and AUC
are taken as performance indicators for the model
[71]. The model includes five dimensions
participation, positivity, frequency, eligibility rate,
and impact [72]. The feasibility analysis of the
selected models is carried out through rigorous
experiments with real data describing the client's
ability to repay loans [73].
In Section II, a systematic literature search
methodology was performed. Here you can see if
the search was performed for the entire year (2018)
onwards specified by the exclusion criteria (Table
4). We obtained a total of 73 journal articles and
conference proceedings. The year with the most
journals published was 2021, with 27 records and
Scopus leads in quantity in this range of years with
25 publications, as shown in Table 8.
Table 8. Publication over the years
Source
Last five years
Total
2018
2019
2020
2021
2022
Taylor &
Francis
1
1
1
3
6
ProQuest
1
5
10
16
WOS
2
2
4
IEEE Xplore
1
1
4
2
8
ScienceDirect
1
1
2
Scopus
1
4
11
7
2
25
EBSCOhost
1
1
2
1
2
7
Wiley Online
Library
2
1
3
IOP
2
2
ERIC
0
Google
Scholar
0
Total
4
9
24
27
9
73
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
68
Volume 10, 2022
Specific machine learning algorithms for
classifying the quality of knowledge in the system
(in this case, a decision tree algorithm using a
training model [58]) work as follows: The algorithm
recognizes whether the knowledge quality is high,
medium, or low [8]. The knowledge quality attribute
[63] is satisfied. This is determined by the training
model itself and must be identified when creating
the training model dataset [17]. Subsequently, the
algorithm analyzes the knowledge. The algorithm
continues with the following perspective when a
perspective receives a score until all views receive a
knowledge quality score [57].
For practitioners, the direct application of the
results provided in the study should be done
carefully using some of the techniques found
(supervised and unsupervised learning). This
evidence suggests that the field of study is
developing promptly, so directly applying the
techniques and tools used in the study will achieve
the desired effect of implementing machine learning
in financial institutions. In particular, we agree with
[26], that there are several methods and theoretical
models to apply data mining technology. We also
believe that careful evaluation of the context in
which the research is conducted is important to
assess the generalizability of the findings to change
in other potential contexts.
4 Conclusion
In conclusion, this study uses a systematic literature
review (SLR). This iterative process combines all
existing literature on a particular topic or research
question. The goal of SLR is to solve a specific
problem by examining and integrating the results of
all state-of-the-art surveys that address two or more
survey questions. The studies found are the best
available in our time on the subject of the study. The
most commonly used criterion is "unsupervised
learning" to determine its basic effectiveness. This
method identifies how far the current research on
the use of machine learning has progressed. The
review work has a total of 73 articles between
journals and congresses. Similarly, it was found that
the most suitable study varied between 16 and 17
pages to answer the highest number of RQs, being
the publication medium Risk, as well as Expert Syst.
Appl. and Math. Probl. Eng. being the journals with
the highest production in the area of machine
learning. It is claimed that RQ1 provides the basis
for future comments on the applicability of personal
credit assessments. Here are some machine learning
methods to show how machine learning can be
applied to an individual's credit score. This method
identifies how far current research on machine
learning has progressed. Future work will be
devoted to improving the literature by also taking
into account questions that have answers by adding
a wider variety of configurations.
References:
[1] P. M. Addo, D. Guegan, and B. Hassani,
“Credit risk analysis using machine and deep
learning models,” Risks, vol. 6, no. 2, 2018,
DOI: 10.3390/risks6020038.
[2] V. Aithal and R. D. Jathanna, “Credit risk
assessment using machine learning
techniques,” Int. J. Innov. Technol. Explor.
Eng., vol. 9, no. 1, pp. 3482–3486, 2019, DOI:
10.35940/ijitee.A4936.119119.
[3] S. Akkoç, “Exploring the nature of credit
scoring: A neuro fuzzy approach,” Fuzzy Econ.
Rev., vol. 24, no. 1, pp. 3–24, 2019, DOI:
10.25102/fer.2019.01.01.
[4] M. Ala’raj, M. F. Abbod, M. Majdalawieh, and
L. Jum’a, “A deep learning model for
behavioural credit scoring in banks,” Neural
Comput. Appl., vol. 34, no. 8, pp. 5839–5866,
2022, DOI: 10.1007/s00521-021-06695-z.
[5] G. S. Alzamora, M. R. Aceituno-Rojo, and H.
I. Condori-Alejo, “An Assertive Machine
Learning Model for Rural Micro Credit
Assessment in Peru,” Procedia Comput. Sci.,
vol. 202, pp. 301–306, 2022, DOI:
https://doi.org/10.1016/j.procs.2022.04.040.
[6] A. Ampountolas, T. N. Nde, P. Date, and C.
Constantinescu, “A machine learning approach
for micro-credit scoring,” Risks, vol. 9, no. 3,
2021, DOI: 10.3390/risks9030050.
[7] A. Ashofteh and J. M. Bravo, “A conservative
approach for online credit scoring,” Expert
Syst. Appl., vol. 176, 2021, DOI:
10.1016/j.eswa.2021.114835.
[8] U. Bhuvaneswari and S. Sophia, “A predictive
analysis of credit risk evaluation and the quality
decision making using different predictive
models,” Int. J. Recent Technol. Eng., vol. 8,
no. 1, pp. 1965–1969, 2019, [Online].
Available:
https://www.scopus.com/inward/record.uri?eid
=2-s2.0-
85068001945&partnerID=40&md5=63ad0475
a06fe201cfa9d754158756c7.
[9] Z. Boz, D. Gunnec, S. I. Birbil, and M. K.
Öztürk, “Reassessment and Monitoring of Loan
Applications with Machine Learning,” Appl.
Artif. Intell., vol. 32, no. 9–10, pp. 939–955,
2018, DOI: 10.1080/08839514.2018.1525517.
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
69
Volume 10, 2022
[10] . L. Breeden, “A survey of machine learning in
credit risk,” J. Credit Risk, vol. 17, no. 3, pp.
1–62, 2021, DOI: 10.21314/JCR.2021.008.
[11] M. Bücker, G. Szepannek, A. Gosiewska, and
P. Biecek, “Transparency, auditability, and
explainability of machine learning models in
credit scoring,” J. Oper. Res. Soc., vol. 73, no.
1, pp. 70–90, 2022, DOI:
10.1080/01605682.2021.1922098.
[12] M. Chen and X. Ma, “An Optimized BP Neural
Network Model and Its Application in the
Credit Evaluation of Venture Loans.,” Comput.
Intell. Neurosci., pp. 1–9, 2022, [Online].
Available:
https://search.ebscohost.com/login.aspx?direct
=true&db=a9h&AN=156652121&lang=es&sit
e=eds-live.
[13] A. Chopra and P. Bhilare, “Application of
Ensemble Models in Credit Scoring Models,”
Bus. Perspect. Res., vol. 6, no. 2, pp. 129–141,
2018, DOI: 10.1177/2278533718765531.
[14] A. Coşer, M. M. Maer-Matei, and C. Albu,
“Predictive models for loan default risk
assessment,” Econ. Comput. Econ. Cybern.
Stud. Res., vol. 53, no. 2, pp. 149–165, 2019,
DOI: 10.24818/18423264/53.2.19.09.
[15] J. R. de Castro Vieira, F. Barboza, V. A.
Sobreiro, and H. Kimura, “Machine learning
models for credit analysis improvements:
Predicting low-income families’ default,” Appl.
Soft Comput. J., vol. 83, 2019, DOI:
10.1016/j.asoc.2019.105640.
[16] D. A. V. de Paula, R. Artes, F. Ayres, and A.
M. A. F. Minardi, “Estimating credit and profit
scoring of a Brazilian credit union with logistic
regression and machine-learning techniques,”
RAUSP Manag. J., vol. 54, no. 3, pp. 321–336,
2019, DOI: https://doi.org/10.1108/RAUSP-03-
2018-0003.
[17] B. Dushimimana, Y. Wambui, T. Lubega, and
P. E. McSharry, “Use of Machine Learning
Techniques to Create a Credit Score Model for
Airtime Loans,” J. Risk Financ. Manag., vol.
13, no. 8, p. 180, 2020, DOI:
https://doi.org/10.3390/jrfm13080180.
[18] A. Ehrhardt, C. Biernacki, V. Vandewalle, P.
Heinrich, and S. Beben, “Reject inference
methods in credit scoring,” J. Appl. Stat., vol.
48, no. 13–15, pp. 2734–2754, 2021, DOI:
10.1080/02664763.2021.1929090.
[19] W. Fengpei, S. Xiang, S. O. Young, and W.
Zhiying, “Personal Credit Risk Evaluation
Model of P2P Online Lending Based on AHP,”
Symmetry (Basel)., vol. 13, no. 1, p. 83, 2021,
DOI: https://doi.org/10.3390/sym13010083.
[20] A. Filchenkov, N. Khanzhina, A. Tsai, and I.
Smetannikov, “Regularization of autoencoders
for bank client profiling based on financial
transactions,” Risks, vol. 9, no. 3, 2021, DOI:
10.3390/risks9030054.
[21] H. Gao, H. Liu, H. Ma, C. Ye, and M. Zhan,
“Network-aware credit scoring system for
telecom subscribers using machine learning and
network analysis,” Asia Pacific J. Mark.
Logist., vol. 34, no. 5, pp. 1010–1030, 2022,
DOI: 10.1108/APJML-12-2020-0872.
[22] A. Gicić and A. Subasi, “Credit scoring for a
microcredit data set using the synthetic
minority oversampling technique and ensemble
classifiers,” Expert Syst., vol. 36, no. 2, 2019,
DOI: 10.1111/exsy.12363.
[23] W. Gui, L. Wang, H. Wu, X. Jian, D. Li, and
N. Huang, “Multiple psychological
characteristics predict housing mortgage loan
behavior: A holistic model based on machine
learning,” PsyCh J., vol. 11, no. 2, pp. 263–
274, 2022, DOI:
https://doi.org/10.1002/pchj.521.
[24] B. R. Gunnarsson, S. vanden Broucke, B.
Baesens, M. Óskarsdóttir, and W. Lemahieu,
“Deep learning for credit scoring: Do or
don’t?,” Eur. J. Oper. Res., vol. 295, no. 1, pp.
292–305, 2021, DOI:
10.1016/j.ejor.2021.03.006.
[25] M. Hanafy and R. Ming, “Classification of the
Insureds Using Integrated Machine Learning
Algorithms: A Comparative Study,” Appl.
Artif. Intell., vol. 0, no. 0, pp. 1–32, 2022, DOI:
10.1080/08839514.2021.2020489.
[26] J. Hu, C. Chen, and K. Zhu, “Application of
Data Mining Combined with Fuzzy
Multicriteria Decision-Making in Credit Risk
Assessment from Legal Service Companies,”
Math. Probl. Eng., vol. 2021, 2021, DOI:
10.1155/2021/2499948.
[27] K. Iwai, M. Akiyoshi, and T. Hamagami,
“Bayesian Network Oriented Transfer Learning
Method for Credit Scoring Model,” IEEJ
Trans. Electr. Electron. Eng., vol. 16, no. 9, pp.
1195–1202, 2021, DOI:
https://doi.org/10.1002/tee.23417.
[28] A. Kim and S.-B. Cho, “An ensemble semi-
supervised learning method for predicting
defaults in social lending,” Eng. Appl. Artif.
Intell., vol. 81, pp. 193–199, 2019, DOI:
10.1016/j.engappai.2019.02.014.
[29] A. Kumar, S. Sharma, and M. Mahdavi,
“Machine Learning (ML) Technologies for
Digital Credit Scoring in Rural Finance: A
Literature Review,” Risks, vol. 9, no. 11, p.
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
70
Volume 10, 2022
192, 2021, DOI:
https://doi.org/10.3390/risks9110192.
[30] S. Lahmiri and S. Bekiros, “Can machine
learning approaches predict corporate
bankruptcy? Evidence from a qualitative
experimental design,” Quant. Financ., vol. 19,
no. 9, pp. 1569–1577, 2019, DOI:
10.1080/14697688.2019.1588468.
[31] J. W. Lee, W. K. Lee, and S. Y. Sohn, “Graph
convolutional network-based credit default
prediction utilizing three types of virtual
distances among borrowers,” Expert Syst.
Appl., vol. 168, 2021, DOI:
10.1016/j.eswa.2020.114411.
[32] H. Li, Y. Cao, S. Li, J. Zhao, and Y. Sun,
“XGBoost Model and Its Application to
Personal Credit Evaluation,” IEEE Intell. Syst.,
vol. 35, no. 3, pp. 52–61, May 2020, DOI:
10.1109/MIS.2020.2972533.
[33] X. Li and Y. Sun, “Application of RBF neural
network optimal segmentation algorithm in
credit rating,” Neural Comput. Appl., 2020,
DOI: 10.1007/s00521-020-04958-9.
[34] L. Liu, “A Self-Learning BP Neural Network
Assessment Algorithm for Credit Risk of
Commercial Bank,” Wirel. Commun. Mob.
Comput., vol. 2022, 2022, DOI:
10.1155/2022/9650934.
[35] S. Liu, R. Wang, and Y. Han, “Research on
Personal Credit Evaluation Based on Machine
Learning Algorithm,” in 2021 6th International
Symposium on Computer and Information
Processing Technology (ISCIPT), 2021, pp.
48–52, DOI:
10.1109/ISCIPT53667.2021.00016.
[36] Y. Liu, A. Ghandar, and G. Theodoropoulos,
“A Metaheuristic Strategy for Feature Selection
Problems: Application to Credit Risk
Evaluation in Emerging Markets,” in 2019
IEEE Conference on Computational
Intelligence for Financial Engineering
Economics (CIFEr), May 2019, pp. 1–7, DOI:
10.1109/CIFEr.2019.8759117.
[37] V. Moscato, A. Picariello, and G. Sperli, “A
benchmark of machine learning approaches for
credit score prediction,” Expert Syst. Appl., vol.
165, Mar. 2021, DOI:
10.1016/j.eswa.2020.113986.
[38] I. Nica, D. B. Alexandru, S. L. P. Crăciunescu,
and Ștefan Ionescu, “Automated Valuation
Modelling: Analysing Mortgage Behavioural
Life Profile Models Using Machine Learning
Techniques,” Sustainability, vol. 13, no. 9, p.
5162, 2021, DOI:
https://doi.org/10.3390/su13095162.
[39] A. Pérez-Martín, A. Pérez-Torregrosa, A.
Rabasa, and M. Vaca, “Feature selection to
optimize credit banking risk evaluation
decisions for the example of home equity
loans,” Mathematics, vol. 8, no. 11, pp. 1–16,
2020, DOI: 10.3390/math8111971.
[40] J. Qi, R. Yang, and P. Wang, “Application of
explainable machine learning based on
Catboost in credit scoring,” J. Phys. Conf. Ser.,
vol. 1955, no. 1, Jun. 2021, DOI:
https://doi.org/10.1088/1742-
6596/1955/1/012039.
[41] C. Qin, Y. Zhang, F. Bao, C. Zhang, P. Liu,
and P. Liu, “XGBoost optimized by adaptive
particle swarm optimization for credit scoring,”
Math. Probl. Eng., vol. 2021, 2021, DOI:
10.1155/2021/6655510.
[42] C. Rao, M. Liu, M. Goh, and J. Wen, “2-stage
modified random forest model for credit risk
assessment of P2P network lending to ‘Three
Rurals’ borrowers,” Appl. Soft Comput. J., vol.
95, 2020, DOI: 10.1016/j.asoc.2020.106570.
[43] B. Ruyu, H. Mo, and L. Haifeng, “A
Comparison of Credit Rating Classification
Models Based on Spark- Evidence from
Lending-club,” Procedia Comput. Sci., vol.
162, pp. 811–818, 2019, DOI:
https://doi.org/10.1016/j.procs.2019.12.054.
[44] F. Shen, X. Zhao, Z. Li, K. Li, and Z. Meng,
“A novel ensemble classification model based
on neural networks and a classifier optimisation
technique for imbalanced credit risk
evaluation,” Phys. A Stat. Mech. its Appl., vol.
526, 2019, DOI: 10.1016/j.physa.2019.121073.
[45] M. Soui, I. Gasmi, S. Smiti, and K. Ghédira,
“Rule-based credit risk assessment model using
multi-objective evolutionary algorithms,”
Expert Syst. Appl., vol. 126, pp. 144–157,
2019, DOI: 10.1016/j.eswa.2019.01.078.
[46] S. Sreesouthry, A. Ayubkhan, M. M. Rizwan,
D. Lokesh, and K. P. Raj, “Loan Prediction
Using Logistic Regression in Machine
Learning,” Ann. Rom. Soc. Cell Biol., vol. 25,
no. 4, pp. 2790–2794, 2021, [Online].
Available:
https://www.proquest.com/scholarly-
journals/loan-prediction-using-logistic-
regression-machine/docview/2582236651/se-
2?accountid=37408.
[47] G. Teles, J. J. P. C. Rodrigues, K. Saleem, and
S. A. Kozlov, “Classification Methods Applied
to Credit Scoring with Collateral,” IEEE Syst.
J., vol. 14, no. 3, pp. 4557–4566, 2020, DOI:
10.1109/JSYST.2019.2937552.
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
71
Volume 10, 2022
[48] G. Teles, J. J. P. C. Rodrigues, R. A. L. Rabêlo,
and S. A. Kozlov, “Comparative study of
support vector machines and random forests
machine learning algorithms on credit
operation,” Softw. Pract. Exp., vol. 51, no. 12,
pp. 2492–2500, 2021, DOI:
https://doi.org/10.1002/spe.2842.
[49] L. Ting, F. Qinwei, K. Qian, and N. Lei,
“Extreme Learning Machine Based on Firefly
Adaptive Flower Pollination Algorithm
Optimization,” Processes, vol. 8, no. 12, p.
1583, 2020, DOI:
https://doi.org/10.3390/pr8121583.
[50] D. Tripathi, D. R. Edla, V. Kuppili, and A.
Bablani, “Evolutionary Extreme Learning
Machine with novel activation function for
credit scoring,” Eng. Appl. Artif. Intell., vol. 96,
2020, DOI: 10.1016/j.engappai.2020.103980.
[51] de C. S. P. Vitor and T. L. C. Bambirra,
“Extreme Wavelet Fast Learning Machine for
Evaluation of the Default Profile on Financial
Transactions,” Comput. Econ., vol. 57, no. 4,
pp. 1263–1285, Apr. 2021, DOI:
https://doi.org/10.1007/s10614-020-10018-0.
[52] C. Wang, Z. Hu, R. Chiong, S. Dhakal, Y.
Chen, and Y. Bao, “A PSO-Based Ensemble
Model for Peer-to-Peer Credit Scoring,” in
2018 14th International Conference on Natural
Computation, Fuzzy Systems and Knowledge
Discovery (ICNC-FSKD), 2018, pp. 412–418,
DOI: 10.1109/FSKD.2018.8687154.
[53] D. Wang, Z. Wu, and B. Zhu, “Controlling
Shareholder Characteristics and Corporate Debt
Default Risk: Evidence Based on Machine
Learning,” Emerg. Mark. Financ. Trade, vol. 0,
no. 0, pp. 1–16, 2022, DOI:
10.1080/1540496X.2022.2037416.
[54] H. Wang, G. Kou, and Y. Peng, “Multi-class
misclassification cost matrix for credit ratings
in peer-to-peer lending,” J. Oper. Res. Soc.,
vol. 72, no. 4, pp. 923–934, 2021, DOI:
10.1080/01605682.2019.1705193.
[55] Z. Wang, C. Jiang, H. Zhao, and Y. Ding,
“Mining Semantic Soft Factors for Credit Risk
Evaluation in Peer-to-Peer Lending,” J. Manag.
Inf. Syst., vol. 37, no. 1, pp. 282–308, 2020,
DOI: 10.1080/07421222.2019.1705513.
[56] C.-H. Weng and C.-K. Huang, “A Hybrid
Machine Learning Model for Credit Approval,”
Appl. Artif. Intell., vol. 35, no. 15, pp. 1439–
1465, 2021, DOI:
10.1080/08839514.2021.1982475.
[57] Y. Wu and Y. Pan, “Application Analysis of
Credit Scoring of Financial Institutions Based
on Machine Learning Model,” Complexity, vol.
2021, 2021, DOI:
https://doi.org/10.1155/2021/9222617.
[58] Y. Xia, L. He, Y. Li, Y. Fu, and Y. Xu, “A
dynamic credit scoring model based on survival
gradient boosting decision tree approach,”
Technol. Econ. Dev. Econ., vol. 27, no. 1, pp.
96–119, 2021, DOI: 10.3846/tede.2020.13997.
[59] D. Xu, X. Zhang, J. Hu, and J. Chen, “A Novel
Ensemble Credit Scoring Model Based on
Extreme Learning Machine and Generalized
Fuzzy Soft Sets,” Math. Probl. Eng., vol. 2020,
2020, DOI: 10.1155/2020/7504764.
[60] J. Xu, L. Zekai, and Y. Xie, “Loan default
prediction of Chinese P2P market: a machine
learning methodology,” Sci. Reports (Nature
Publ. Group), vol. 11, no. 1, 2021, DOI:
https://doi.org/10.1038/s41598-021-98361-6.
[61] Y.-Z. Xu, J.-L. Zhang, Y. Hua, and L.-Y.
Wang, “Dynamic credit risk evaluation method
for e-commerce sellers based on a hybrid
artificial intelligence model,” Sustain., vol. 11,
no. 19, 2019, DOI: 10.3390/su11195521.
[62] G. Yang, “Research on Financial Credit
Evaluation and Early Warning System of
Internet of Things Driven by Computer-Aided
Technology,” Comput. Aided. Des. Appl., vol.
19, no. S6, pp. 158–169, 2022, DOI:
10.14733/cadaps.2022.S6.158-169.
[63] J. You, G. Li, and H. Wang, “Credit Grade
Prediction Based on Decision Tree Model,” in
2021 16th International Conference on
Intelligent Systems and Knowledge
Engineering (ISKE), Nov. 2021, pp. 668–673,
DOI: 10.1109/ISKE54062.2021.9755326.
[64] X. Yu, Q. Yang, R. Wang, R. Fang, and M.
Deng, “Data Cleaning for Personal Credit
Scoring by Utilizing Social Media Data: An
Empirical Study,” IEEE Intell. Syst., vol. 35,
no. 2, pp. 7–15, 2020, DOI:
10.1109/MIS.2020.2972214.
[65] Y. Yuan, X. Gong, M. Guo, Y. Han, and H. Li,
“Research on Personal Credit Evaluation of
Commercial Banks Under Ensemble Learning
Framework,” in 2020 2nd International
Conference on Applied Machine Learning
(ICAML), Oct. 2020, pp. 29–38, DOI:
10.1109/ICAML51583.2020.00015.
[66] Z. Yuhao, “Research on personal credit
evaluation of internet finance based on
blockchain and decision tree algorithm,”
EURASIP J. Wirel. Commun. Netw., vol. 2020,
no. 1, Dec. 2020, DOI:
https://doi.org/10.1186/s13638-020-01819-w.
[67] L. Zhang and J. Fan, “Formal Credit-Assisted
New Agricultural Business A Multifactor
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
72
Volume 10, 2022
Analysis Based on BP Neural Network.,” Mob.
Inf. Syst., pp. 1–11, 2022, [Online]. Available:
httpssearch.ebscohost.comlogin.aspxdirect=tru
e&db=iih&AN=156288423&lang=es&site=eds
-live.
[68] S. Zhang, Y. Hu, and Z. Tan, “Research on
borrower’s credit classification of P2P network
loan based on LightGBM algorithm,” Int. J.
Embed. Syst., vol. 11, no. 5, pp. 602–612, 2019,
DOI: 10.1504/IJES.2019.102435.
[69] T. ZHANG, W. ZHANG, W. XU, and H.
HAO, “Multiple instance learning for credit
risk assessment with transaction data,”
Knowledge-Based Syst., vol. 161, pp. 65–77,
2018, DOI: 10.1016/j.knosys.2018.07.030.
[70] Y. Zhao, “Research on Personal Credit
Evaluation of Internet Finance Based on
Blockchain and Decision Tree Algorithm,”
EURASIP Journal on Wireless
Communications and Networking. Research
Square, Durham, Aug. 13, 2020, DOI:
https://doi.org/10.21203/rs.3.rs-34146/v2.
[71] X. Zhou, W. Zhang, and Y. Jiang, “Personal
credit default prediction model based on
convolution neural network,” Math. Probl.
Eng., vol. 2020, 2020, DOI:
10.1155/2020/5608392.
[72] J. Zhu et al., “A Deep Learning-Oriented
Personal Credit Model for Online Learning
Platform,” in 2020 15th International
Conference on Computer Science Education
(ICCSE), 2020, pp. 838–843, DOI:
10.1109/ICCSE49874.2020.9202397.
[73] P. Ziemba, J. Becker, A. Becker, A. Radomska-
Zalas, M. Pawluk, and D. Wierzba, “Credit
Decision Support Based on Real Set of Cash
Loans Using Integrated Machine Learning
Algorithms,” Electronics, vol. 10, no. 17, p.
2099, 2021, DOI:
https://doi.org/10.3390/electronics10172099.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
WSEAS TRANSACTIONS on COMPUTER RESEARCH
DOI: 10.37394/232018.2022.10.9
Cano Chuqui Jorge, Ogosi Auqui José Antonio,
Guadalupe Mori Victor Hugo,
Obando Pacheco David Hugo
E-ISSN: 2415-1521
73
Volume 10, 2022