You are on page 1of 18

Sustainable and Resilient Infrastructure

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tsri20

Resilient and sustainable supplier selection: an


integration of SCOR 4.0 and machine learning
approach

Md Muzahid Khan, Imranul Bashar, Golam Morshed Minhaj, Absar Ishraq


Wasi & Niamat Ullah Ibne Hossain

To cite this article: Md Muzahid Khan, Imranul Bashar, Golam Morshed Minhaj, Absar Ishraq
Wasi & Niamat Ullah Ibne Hossain (2023): Resilient and sustainable supplier selection:
an integration of SCOR 4.0 and machine learning approach, Sustainable and Resilient
Infrastructure, DOI: 10.1080/23789689.2023.2165782

To link to this article: https://doi.org/10.1080/23789689.2023.2165782

© 2023 The Author(s). Published by Informa


UK Limited, trading as Taylor & Francis
Group.

Published online: 16 Jan 2023.

Submit your article to this journal

Article views: 1266

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=tsri20
SUSTAINABLE AND RESILIENT INFRASTRUCTURE
https://doi.org/10.1080/23789689.2023.2165782

ARTICLE

Resilient and sustainable supplier selection: an integration of SCOR 4.0 and


machine learning approach
Md Muzahid Khana, Imranul Bashara, Golam Morshed Minhaja, Absar Ishraq Wasia
and Niamat Ullah Ibne Hossain b
a
Department of Industrial and Production Engineering, Military Institute of Science and Technology, Dhaka, Bangladesh; bEngineering
Management Department, College of Engineering and Computer Science, Arkansas State University, Jonesboro, Arkansas, USA

ABSTRACT ARTICLE HISTORY


The purpose of this research paper is to implement a machine learning model with the integration Received 8 September 2022
of the supply chain occupational reference (SCOR) model to develop an artificial intelligence-based Accepted 29 December 2022
system for resilient and sustainable supplier selection for a pharmaceutical company. Initially, the KEYWORDS
SCOR 4.0 model with the integration of Best Worst Method (BWM) has been used to develop the Supplier selection; resilience;
framework of customer satisfaction and to identify the critical elements of the suppliers. Later, the machine learning; SCOR 4.0;
gradient boosting machine learning model has been implemented to classify the supplier as well Pharmaceutical Supply
as rank the suppliers from best to worst based on the acceptability score. The result shows that the Chain; sustainability
gradient boosting algorithm performs well as a classifier, where the supplier with the most
acceptability score represents the best supplier and the supplier with the least acceptability
score represents the worst supplier. This study contributes to our understanding of how and
when integrated SCOR and machine learning models can help improve supplier selection.

1. Introduction
a competitive advantage and improve organizational per­
Suppliers are an important component of the supply formance. It establishes a common language for deciding,
chain because their performance has an indirect and organizing, and implementing supply chain procedures
significant impact on customer satisfaction. Over the (Delipinar & Kocaoglu, 2016; Munson & Santa-Eulalia,
last two decades, resilience and sustainability concerns 2014). The SCOR model has been successfully used in
have grown in importance in SCM (Ghamari et al., 2021). a variety of sectors, including the pharmaceutical industry,
The idea of sustainability has emerged as an essential way to identify efficient supplier solutions and this model can
of thinking across a variety of business domains as be used in conjunction with other Multi-Criteria Decision
a direct result of the rising awareness of social occupa­ Making (MCDM) tools and methods to facilitate supplier
tions and the preservation of the natural environment. selection (Es-Satty et al., 2020). Nevertheless, in today’s
Apart from that, the company supply chain is increas­ digital era, there are fundamental changes in the structure
ingly threatened by pandemics, natural, man-made, or of supply chains such as the integration of new technology.
mechanical hazards such as surges, seismic tremors, fires, Supply chain structure development with evolving tech­
transportation mishaps, strikes, attacks, and so on. These nologies is critical for staying up with the times. The new
disasters interrupt the supply chain, which harms firms supply chain requirements should be precisely determined,
through lost productivity, revenue, competitive advan­ and new metrics should be carefully determined, taking
tage, profitability, and so on. As a consequence of this, into account new requirements. Digital technology and
it is necessary to take a resilient approach to the supply information systems are newly treated as two new metrics
chain to secure the buyer from disruptions and deficien­ in the SCOR model, and this novel model is known as
cies. Since a supplier can affect the overall performance of SCOR 4.0, which is a structured three-level hierarchical
a supply chain, companies must take resilience in supplier framework to analyze supply chain performance (Ayyildiz
selection decisions to reduce risk (Amindoust, 2018; & Taskin Gumus, 2021). Previously researchers have
Didier et al., 2018; Ghamari et al., 2021; Zhaoa et al., widely used different MCDM methods such as AHP,
2021). Hybrid Best Worst Method (BWM), Fuzzy Topsis,
The SCOR model was established by the Supply Chain Pythagorean fuzzy AHP, or other decision-making models
Council (SCC) to monitor chain performance to acquire with the traditional SCOR model for supplier selection

CONTACT Niamat Ullah Ibne Hossain nibnehossain@astate.edu


© 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
2 M. D. M. KHAN ET AL.

(Ayyildiz & Taskin Gumus, 2021; Huang et al., 2021; Lima- A company’s performance is heavily dependent on the
Junior & Carpinetti, 2016; Nazim & Yaacob, 2017). performance of its suppliers. For this, the evaluation of
Nevertheless, MCDM procedures frequently result in a supplier’s performance for development should be
enormously complicated models that are challenging to linked with the entire supply chain’s performance eva­
handle and analyze. luation. The SCOR model is one way of performance
Machine learning (ML) has recently attracted more measurement of the supply chain that is frequently used
attention as a result of its effectiveness in a variety of by practitioners. The Supply Chain Council (SCC)
applications ranging from image categorization to established this model back in 1996, which provides
a variety of decision-making challenges. The applications a hierarchy of performance evaluation criteria (Lima-
of ML approaches have increased dramatically in recent Junior & Carpinetti, 2016). Hwang et al. (2008) created
years due to the explosion of data. In a predictive setting, a SCOR model based on three-level performance
methods such as data envelopment analysis supervised metrics to evaluate the supply chain process and
learning, and unsupervised learning has shown very out­ enhance the performance of sourcing. The researchers
standing performance (Tavana et al., 2016). Furthermore, used a regression model to examine the performance
machine learning can tolerate inaccuracies, uncertainty, metrics. Raut et al. (2012) suggested a hierarchical
and imprecise information to achieve robustness when approach based that incorporates AHP and Data
replicating human decision-making behavior. These func­ Envelopment Analysis (DEA) and considers both the
tions not only solve the problem of scalability and rapidity, quantitative and qualitative factors for supplier selec­
but they also reduce the drawbacks of earlier approaches tion. Authors suggested that the most likely candidates
and meet the demands of ever more difficult supplier net­ for such programs are the highly efficient and low effi­
works (Sepehri, 2020). The developments of artificial intel­ cient suppliers. Badi and Ballem (2018) used BWM for
ligence (AI) have drawn the attention of many researchers, supplier selection in the pharmaceutical supply chain in
considering how they can be used in supplier risk manage­ Libya. The researchers considered five different criteria
ment (Abdulla et al., 2019; Batta, 2020; Mirkouei & such as cost, quality, supplier profile, flexibility, and
Haapala, 2014). However, as the amount of data that delivery for supplier selection. The study’s results
needs to be processed expands the efficiency of the tradi­ showed that price is the most important factor, followed
tional methods decreases. Because of this, researchers came by quality in second place and the company’s profile in
up with several approaches such as machine learning, big third. Along the same line, Oroojeni Mohammad Javad
data analysis, and cloud computing that have a high cap­ et al. (2020) implemented BWM to find the green sup­
ability to analyze and interpret enormous amounts of data. plier Khouzestan Steel Company located in Iran. The
To give this research a new dimension, we implemented researchers proposed a three-step method that helps the
machine learning with the integration of SCOR 4.0 company decision-makers rank different suppliers and
implemented. choose the one with the best green innovations abilities.
The object of this research is to develop a machine With the right deployment of resources, these ven­
learning model-based supplier selection system with the dors with efficient infrastructure could achieve high-
integration of the SCOR 4.0 model. First, we developed performance levels. In another research, Lu et al.
a 3-level based SCOR 4.0 model to identify the supplier (2016) created a framework using the SCOR model for
performance related criterion and attributes for a local evaluating the performance of humanitarian supply
pharmaceutical company. Secondly, To calculate the chains. For performance evaluation, the researchers
weights of each attribute and criterion, BWM has been developed a 14 metrics-based SCOR model addition to
used for all the 3 levels of the SCOR 4.0 model. Finally, we this, numerous decision-making strategies for dealing
implement a gradient boosting machine learning algo­ with the process of supplier evaluation, particularly in
rithm to classify the supplier and find out the most poten­ the context of supplier selection, have been presented in
tial supplier. the relevant body of academic literature (Ayyildiz &
Taskin Gumus, 2021; Luitzen et al., 2001; Wendler-
Bosco & Nicholson, 2020; Wu & Barnes, 2018). Mañay
2. Literature review et al. (2022) used the SCOR model with the integration
of MCDM to assess the supply chain performance of the
2.1. SCOR model for supply chain performance
Ecuadorian flower industry. The authors measure the
evaluation
performance level of each construct and identified the
Identifying relevant and appropriate criteria for supplier external factors that need to be improved. Jain et al.
performance assessment is one of the most challenging (2022) used the SCOR model as the basic performance
parts of supplier selection (Alavi et al., 2021). evaluation hierarchy. In this research, the authors
SUSTAINABLE AND RESILIENT INFRASTRUCTURE 3

assessed the sustainability performance of a company’s Regressor Chain (RRC) method for better forecasting
e-waste supply chain to meet the sustainability goal. In accuracy and identification of suitable suppliers and
another research, Huang et al. (2021) applied Data DEA order quantities. In another research, Ayyildiz and
and SCOR model to evaluate the operational efficiency Taskin Gumus (2021) proposed a three-level based
of ports. The authors analyzed the operational efficiency SCOR 4.0 model for the evaluation of supply chain
of different key ports of different cities based on the performance. The researchers integrated hybrid BMW
ship’s inward and outward. Along the same line, He and and Pythagorean fuzzy AHP with the SCOR 4.0 model
Zhu (2022) developed a risk indicator system using to find out the weights of the metrics. Alavi et al. (2021)
DEA and SCOR model for the elderly care service proposed a model based on MCDM and machine learn­
industry to enhance risk management capabilities. ing for sustainable supplier selection under uncertainty.
The proposed decision support system can be used for
supplier risk assessment to manage uncertainties in the
2.2. Machine learning for supplier selection
field of health care, manufacturing, or other sectors.
In recent years, a variety of strategies have emerged, Zhang et al. (2016) investigated the Ranking Neural
each based on a different set of criteria, to identify the Network (RankNet) ML model for supplier evaluation
best possible suppliers. The MCDM approach is one of and selection as well. In their research, new suppliers
the methods that is utilized the most frequently. were taken into account for each customer order, even
However, the drawback of existing multi-criteria sup­ when the customer placed multiple orders. The ML
plier selection and performance evaluation models for model that was developed as a result of this research
development is the model itself. Methods such as the improved with time because it was able to acquire
Analytic Hierarchy Process (AHP), the Analytic knowledge from previous experiences, and the more
Network Process (ANP), and the Fuzzy Analytic data that was available, the simpler it became. A Multi-
Hierarchy Process (Fuzzy AHP) are used in the majority Agent Systems (MASs) method has been proposed by
of the studies and these methods require decision- Ghadimi et al. (2019) to address sustainable supplier
makers to make judgments based on comparisons. performance evaluation and selection procedures.
Even though these methods are good for dealing with These processes aim to establish an appropriate struc­
vague or qualitative information, they often restrict the ture of communication channels, organized exchange of
number of factors and suppliers that can be examined at information, and clarity among suppliers and manufac­
analyzed the same time (Park et al., 2010; Pilkington & turers. In another research, Mustafa et al. (2020) imple­
Mahmoud, 2016; Zeydan et al., 2011; Zhao et al., 2021). mented several ML models such as artificial neural
With the rise of AI, ML has been used in several ways in network (ANN), DT classifier, bayesian classifiers, and
various domains of supply chain management, such as k-nearest neighbor (KNN). They developed a data set
selecting, evaluating, and developing resilient and sus­ consisting of 1000 supplier profiles and classify the data
tainable suppliers. Mirkouei and Haapala (2014) inves­ using these ML models to identify the risky suppliers.
tigated how to choose the most suitable supplier for Along the same line, Sepehri (2020) This study pro­
a biomass supply chain by taking into account both posed a framework for predicting supplier future value
qualitative and quantitative aspects. The researchers that incorporates contract history data, interpersonal
implemented Decision Tree (DT) for supplier classifica­ value, and supply network attributes for supplier selec­
tion and Support Vector Machines (SVM) to identify tion and managing relationships. The researchers
the potential supplier. Along the same line, Abdulla et implemented Clustering techniques and DT for supplier
al., (2019) proposed a hybrid supplier selection model performance evaluation, selection, and relationship
by combining AHP and the machine learning model. management.
The researchers applied a DT classifier to classify the The relevant literature on supplier selection used
good and bad suppliers. In another research, Cavalcante in this research is mentioned in Table 1, followed by
et al. (2019) proposed a data-driven model combined existing research gaps and the contribution of this
with a SML model for resilient supplier selection. The research.
findings imply that this model can be useful for identi­ The above discussion shows that despite
fying portfolios of hidden, high-risk suppliers and for a gradually growing body of knowledge on resilient
prioritizing risk mitigation actions. Islam et al. (2021) and sustainable supplier selection, till now no
investigated the influences of forecasting methods on research has used the integration of SCOR 4.0 and
supplier selection. The researchers suggested a model ML approach together that focuses on assessing sup­
with two stages for both the selection of suppliers and plier performance evaluation and selection. To bridge
the distribution of orders. They applied Relational this gap, we adopted the application of SCOR 4.0
4 M. D. M. KHAN ET AL.

and a supervised machine learning approach to rank healthcare services has a significant impact on the
the supplier based on their profile and identify the quality and form of patient care and accounts for
most potential suppliers. The adoption of SCOR 4.0 a significant proportion of supply chain expenses
and gradient boosting machine learning model that (Ishizaka et al., 2022). Apart from that, the
could rank the supplier based on their profile and COVID-19 pandemic’s severe shortage of essential
identify the most potential supplier. The contribution medical supplies created significant procurement dif­
of this study is summarized as follows: ficulties in the healthcare supply chain. The discus­
sion on how to improve supply chain resilience in
● Integration SCOR 4.0, and machine learning healthcare has picked up steam because maintaining
approach to select the best resilient and sustainable the availability of such items during disruptions is
supplier. essential. Healthcare supply chains now have more
● Implementation of SCOR 4.0 to develop the assess­ interdependencies and efficiency awareness due to
ment framework of supply chain performance eva­ expanding demand, rising costs, and more competi­
luation and customer satisfaction. tion. As a supply network’s complexity rises, risk also
● Implementation of a supervised machine learning does as well (Craighead et al., 2020; Hossain et al.,
model (gradient boosting) to rank the suppliers 2016). Hospitals across the world have been
and find out the best supplier. impacted by the severe shortage of essential medical
● Demonstrate the efficacy of SCOR 4.0 integrated supplies like personal protective equipment, ventila­
ML approach in supply chain management. tion systems, and medicines for COVID-19 treat­
ment (Chamola et al., 2020). During this pandemic,
supply chain resilience is significantly impacted by
3. Case study: pharmaceutical industry
the purchasing decisions made by pharma companies
In the pharmaceutical sector, the challenge of sus­ and medical supply manufacturers, as well as by
tainable supplier selection is crucial. Global business suppliers (Spieske et al., 2022). Considering these
organizations have become increasingly concerned concerns, we have chosen the pharmaceutical indus­
about the environmental impact of their manufactur­ try as our research field.
ing and other practices during the past few decades.
Thus, sustainability factors are integrated into the
supplier selection process to create a green supplier 4. Methodology of research
selection. This issue is enhanced by the fact that the
pharmaceutical sector is unique from other indus­ This section describes the detailed research methodol­
tries in that patient care is its primary objective ogy of our approach. This research was conducted in
(Stević et al., 2020). Managing and selecting pharma­ a structured way with several stages. The detailed
ceutical raw materials, and supplies for vital research methodology is presented in the following

Table 1. Summary of relevant themes in supplier selection literature.


Authors Approach Area
Hwang et al. (2008) Level 3 SCOR model Sourcing performance improvement
Raut et al. (2012) AHP and DEA Highly efficient and low efficient supplier identification.
Lu et al. (2016) 14 metrics-based SCOR model Performance evaluation of humanitarian supply chains
Mañay et al. (2022) AHP and SCOR model Supply chain performance of the floricultural
sector
Jain et al. (2022) MCDM and SCOR model Sustainable supply chain
Huang et al. (2021) Data envelopment analysis and SCOR model Operational efficiency evaluation of ports
He and Zhu (2022) Data envelopment analysis and SCOR model Elderly services supply chain
Mirkouei and Haapala (2014) DT and SVM Supplier classification of a biomass supply chain network
(Amindoust, 2018) Hybrid intelligent method Resilient and sustainable supplier selection
Abdulla et al., (2019) AHP and DT Supplier classification
Cavalcante et al. (2019) Data driven model Identification of high-risk suppliers
(Amiri et al., 2020) Fuzzy BWM Sustainable supplier selection
Islam et al. (2021) Relational Regressor Chain (RRC) Supplier selection and order allocation.
Ayyildiz and Taskin Gumus (2021) Hybrid BMW and Pythagorean fuzzy AHP with SCOR 4.0 Supply chain evaluation
Alavi et al. (2021) MCDM and Machine learning Supplier risk assessment
(Zhaoa et al., 2021) IVIKOR Resilient and sustainable supplier selection
(Ghamari et al., 2021) MICMAC and MCDM Resilient and sustainable supplier selection
(Zhang et al., 2016) RankNet Supplier selection
(Ghadimi et al., 2019) Multi-agent systems Intelligent sustainable supplier selection
(Mustafa et al., 2020) ANN, DT, KNN The risk profile of supplier’s identification
(Sepehri, 2020) Clustering and DT Supplier selection and relationship
SUSTAINABLE AND RESILIENT INFRASTRUCTURE 5

Figure 1 followed by a detailed description of each selected from a wide range of educational backgrounds
phase. and with different levels of managerial and industry
experience to ensure that we take into account every
individual’s point of view in the selection process. This
4.1. Phase I: development of the SCOR 4.0 model
resulted in a reduction in bias and was sufficient for the
This section discusses the development strategies of the purpose. Through their years of experience, the experts
SCOR 4.0 model for our research. in Table 2 are capable of figuring out which factors are
most important when choosing a supplier.
4.1.1. Data collection and analysis Figure 2 illustrates the strategy for the data collection
Data collection and analysis include identification, and analysis process. We collected data from different
screening, eligibility, and final selection. The findings local pharmaceutical companies, and academic and
were analyzed using a combination of theory-driven industry experts. Initially, we received a screened data
and data-driven approaches. For our data collection, set of 33 suppliers from different local suppliers of
a survey and interview were conducted with different a pharmaceutical company. We screened all the data
local industry experts, academicians, and researchers via sets and excluded the data which were not closely
face-to-face and phone calls interviews. However, related to this research. In the next step, we evaluated
regarding the inclusion and exclusion criteria, industry our existing data to check eligibility. We validated the
experts work in the supply chain area for pharmaceu­ data by fitting it into a normal distribution. Finally, we
tical companies, while academic experts conduct exten­ had a data set of 28 suppliers which were eligible for this
sive research in this area are chosen. As a result, these research.
experts are capable of understanding supplier data, eva­
luation criteria, and selection criteria effectively. 4.1.2. Data validation
A paper-based survey form was sent out to ten experts, After data collection, we validated the data using the
of which one was unwilling to participate, and three normal distribution. In this phase, we evaluated the
have not completed the entire survey. Finally, we pro­ eligibility of the gathered data using MINITAB statisti­
ceed with the survey input from six experts. These cal software and fit in the normal distribution. Here the
experts have direct experience in the area of subject null hypothesis represents that data is normally distrib­
interest as well as decent managerial experience (two uted and the alternative hypothesis represents that data
of them for a period of five years), while others have is not normally distributed. We used the p value from
extensive hands-on experience. Our experts have been the normal distribution of every distinct criterion.

Phase I: Develop SCOR 4.0 Model

• Identify the factors for the SCOR 4.0 model factors based on literature
analysis and expert
• Data collection, analysis and validation
• Create a three level metrics for the pharmacutical company

Phase II: Implementation of Best Worst Method (BWM)

• Asessing the weights of the criterion of each of metrics


• Find out which criteria are the most important (Best) and which are the
least important (Worst)

Phase III: Implementation of Machine Learning Model

• Selection of supervised machine learning model


• Implementation of gradient boosting algorithm

Phase IV: Results Analysis

• Conduct analysis based on the output of machine learning model


• Ranking of the supplier based on the accepatance score

Figure 1. Research methodology.


6 M. D. M. KHAN ET AL.

Table 2. Supply chain expert’s profiles from industry and improve the efficiency of their supply networks and give
academia. a process-based way to measure the performance of the
Level of Managerial Industry experience supply chain (Ayyildiz & Taskin Gumus, 2021). Plan,
Experts education experience (Years) (Years)
Expert 1 PhD 2 5
Source, Make, Deliver, and Return are the five manage­
Expert 2 Masters 7 13 ment processes that makeup SCOR which are further
Expert 3 Masters 5 10 broken down into categories, elements, tasks, and activ­
Expert 4 Masters 1 5
Expert 5 Bachelors 0 3 ities. In addition, performance attributes and metrics
Expert 6 Bachelors 0 2 are the two types of components that comprise the
performance section of the SCOR model.
A performance attribute is a compilation of metrics
A statistically significant result is one with a p-value
that are utilized in the process of describing a strategy,
greater than 0.05, in that case, the null hypothesis is
whereas an attribute is utilized in the process of evalu­
accepted which means the alternative hypothesis is
ating the objectives of a strategy (Ayyildiz & Taskin
rejected. As we have a total of 41 criteria, all p-values
Gumus, 2021). Figure 3 shows the visual representation
were determined individually. We have found all our
of the SCOR model.
p-value to be greater than 0.05, which signifies that the
We identified the performance attributes and devel­
given data is normally distributed.
oped a 3-level based SCOR 4.0 model for the pharma­
ceutical industry from a detailed literature review and
4.1.3. Create SCOR 4.0 model face-to-face interviews with academic and industry
The SCOR model is a screening tool for supply chain experts. Table 3 shows the 3 levels SCOR 4.0 model
performance management. It helps users understand where level 1 defines the process that is used to describe
how an organization runs its business and figure out the high-level configuration in a supply chain and level 2
which parts are most important to better customer differentiates the level 1 process based on strategies and
experience and satisfaction (Munson & Santa-Eulalia, tactics. Finally, the level 3 processes represent the step
2014). In 1996, the Supply Chain Council (SCC) came needed to be performed to accomplish the level 2
up with the idea of the SCOR model to help businesses process.

Figure 2. Strategy for data collection and analysis process.


SUSTAINABLE AND RESILIENT INFRASTRUCTURE 7

Figure 3. Framework of SCOR model (Zangoueinezhad et al., 2011).

4.2. Phase II: implementation of best worst method use it to determine the importance weights of the cri­
(BWM) teria that are employed in the stages of a problem. BWM
can also be applied to the solution of a variety of deci­
This section discusses the detailed implementation pro­
sion-making issues arising in the real world, including
cedure of BWM for this research work.
those of business, healthcare, and engineering. In the­
ory, this strategy can be applied in any situation in
4.2.1. Best worst method (BWM) which the objective is to rank a group of options and
Rezaei (2015) introduced BWM as an MCDM approach pick one of those rankings as the best option (Pamučar
for calculating the weight of each criterion by minimiz­ et al., 2020). In this research, we chose BWM because of
ing the number of pairwise comparisons between cri­ its calculation flexibilities and decision-making
teria and inconsistent preferences. It is also possible to properties.

Figure 4. Supplier selection using machine learning (Cavalcante et al., 2019).


8 M. D. M. KHAN ET AL.

Table 3. SCOR 4.0 model for the pharmaceutical industry.


Level 1 metric Level 2 metric Level 3 metric
Reliability Quality Quality Certifications
Product specification
Defect percentage
Packaging and labeling
Transportation quality
Research, development, and innovation
Accuracy Order fulfillment
Delivery performance
Document accuracy
Flexibility Delivery On-time delivery
Deliver proper quantities Safe Delivery
Responsiveness Cycle time Delivery cycle time
Source cycle time
Delivery retail cycle time
Quantity supplied Supply type
Lead time
Cost Transportation cost Shipping
Material cost Mode of transaction
Price
Financial status Previous record
Digital technology Ability Dynamism
Collaboration
Method Digital marketing
Artificial intelligence
Other methods
Systems Facility
Applicability Consistency
Information systems Integration Integrated database
Integrated interface
Partner integration
Content Data management
Tactical level
Documentation Supplier information
Delivery documentation
Qualification and validation
Certification
Services Customer relationship management Communication process
Tracking management system
After sales service Warranty
Return policy

To determine the weight of the criteria for the ML where Aiw shows the level of importance of criterion
model, we applied BWM to all three levels of the SCOR j over the least important criterion W (Ayyildiz &
4.0 model. The steps for finding the weights of each Taskin Gumus, 2021).
criterion are given below. Step 5: Determine the optimal weight of each of the
Step 1: Determine the set of decision criteria (C1, C2, criteria (W1, W2, W3, . . .. . . . . .Wn). The main objective
C3, . . . . . . Cn). of Step 5 is to find the optimal weights �of the criteria
� to
�WB �
Step 2: Identify the Most Important (Best) and Least maximize
� � the disparity between � Wj A Bj � and
Important (Worst) criteria. � Wj �
�WB AjW � have to be minimized. The following math­
Step 3: Make a pairwise linguistic comparison using
the linguistic scale as shown in Table 1. Table 4 shows ematical problem can be constructed by incorporating
the scales of linguistic terms and constructs the best to nonnegative constraints as well as the constraint for the
another vector, AB = (AB1, AB2, AB3 . . . . ABn) where ABj
represents the level of importance of the most impor­
Table 4. Linguistic scale for best worst method (BWM).
tant criteria B over criterion j (Ayyildiz & Taskin
Numbers Linguistic Term
Gumus, 2021).
1 Equally importance
Step 4: Similarly, to Step 3, make a pairwise compar­ 2 Weakly importance
ison with the worst criterion to the other criterion and 3 Moderately importance
4 Very importance
develop the others to the worst vector. 5 Absolutely importance
6 Strongly importance
7 Strongly plus importance
Aw ¼ ðA1W ; A2W ; . . . . . . . . . :AnW ÞT 8 Very strong importance
9 Extreme importance
SUSTAINABLE AND RESILIENT INFRASTRUCTURE 9

total sum of all weights (Ayyildiz & Taskin Gumus, Table 5. Glossary of level 1 metric.
2021). Abbreviation Elaboration
This mathematical model can be represented as REL Reliability
FLEX Flexibility
Equation (1) as follows: RES Responsiveness
COS Cost
min � DT Digital Technology
IS Information Systems
Subject to SER Service
8 Pn
> W ¼1
>
> � � j
j¼1
>
> �WB � Table 6. Weight of level 1 criterion.
>
< �� Wj ABj �� � � ; for all j
Level 1 Weight
� Wj � (1)
>
> �WB AjW � � � ; for all j REL 0.2608
>
> EX 0.1565
>
> Wj � 0; for all j
: RES 0.1565
COS 0.0782
DT 0.0260
IS 0.0347
SER 0.2869
4.2.2. Calculation of the weights of the criterion � 0.0521
As stated earlier, we applied BWM to find the optimal
weights and rank of the criterion. To rank this type of
large matrix, each level needs to be weighted. In this data to improve their operations. Companies can
study, first, we found out the weights of the level 1 forecast market changes with the help of machine
criterion both from industry and academic experts and learning and can ensure that the forecasts are cor­
reviewed the survey data to find out the importance of rect and updated by constantly receiving new data
each criterion in terms of weightage. After that, we (Judith & Daniel, 2018). Many different algorithms
followed the weight calculation steps of BWM men­ are used in machine learning to improve operations
tioned in section 4.2.1 both for level 2 and level 3, and explain information with possible outputs. The
respectively. We needed at least three criteria in the ML model learns to predict the output using the
MCDM models to find out the weights. But in our training data. After the model has been trained, it
level 2, we had some metrics where the criteria are will respond to specific input by returning the
only 2. In that case, we took expert opinions. Table 5 desired output when that input is given (Allgurin
shows the terms used in level 1. The glossary of level 2 & Filip, 2018). The ML model is broken into three
and level 3 metrics are given in Table A1 in the appendix major parts. The first part is the decision process,
section. where the model inputs some labeled or unlabeled
After the implementation of BWM for level 1 metrics, data (Zhang, 2017). Based on these data, it predicts
We obtained the weights for REL, FLEX, RES, COS, DT, a pattern for their decision-making process.
IS, SER as 0.26087, 0.156522, 0.015622, 0.078261, Followed by a decision process, an error function
0.026087, 0.034783, 0.286957 respectively. We also got helps the model rectify the error and store necessary
the value of �� for level 1 matrices as 0.05217. Table 6 information in the memory to assist in future deci­
represents the weight of the criteria of level 1 metrics. sion-making (Batta, 2020). An example of supplier
Table 7 shows the final weights of the SCOR 4.0 model. selection using machine learning is illustrated in
A ranking was also done to find out the best and most Figure 4.
important criteria in each of the groups of metrics. In this research, we selected the gradient boosting
ML model because it uses advanced regularization that
helps to improve model generalization abilities.
4.3. Phase III: implementation of machine learning
Gradient boosting is a highly effective scalable end-to-
model
end tree boosting system and a widely used ML method.
This section discusses the selection and implementation The concept of boosting came from the sense that
of the ML model in this research. a weak learner can be improved (Bentéjac et al., 2021;
Dorogush et al., 2018). Apart from that, if the iterative
4.3.1. Gradient boosting machine learning (ML) approach is not appropriately regularized, this algo­
model rithm may undergo from over-fitting (Bentéjac et al.,
Machine Learning (ML) is a popular topic right 2021). The attributes of gradient boosting are listed
now for businesses looking for novel ways to use below.
10 M. D. M. KHAN ET AL.

● Learning rate algorithm. This false positive value causes Type-I error
● Depth (maximum) of the tree due to the prediction of the opposite sign datapoint. The
● The rate of subsampling for random samples size model predicted one opposite signed value, whereas there
● Number of features was no false negative (FN) value which means no value
● The minimum of samples needed to segment an was falsely predicted.
internal node
4.4.2. Ranking of the suppliers
As discussed earlier, we applied the gradient boosting
ML algorithm to find out the acceptance score of the
Algorithm 1 Represents the gradient boosting suppliers. To prevent our model from overfitting, we
pseudo code that we have implemented in this divided the train and test dataset into a 70:30 ratio
research. (Roshan et al., 2022). We also applied feature extraction
and feature selection to increase the accuracy of the
model, eliminate redundant data and boost the learning
4.4. Phase IV: results analysis process. Initially, we fitted our supplier data in a data
frame and fitted X and Y with the supplier ratings and
In this section, the experiment results are presented and the ranking of the criteria obtained from BWM shown
discussed with suitable illustrations. in Table 7. We fitted the X and Y training dataset with
the gradient boosting classifier with an ‘n’ estimator of
4.4.1. Confusion matrix of gradient boosting 10 and a random state of 30. After finally fitting the
The confusion matrix visualizes the performance of model, the feature importance was determined between
a classification model. Figure 5 shows a 2 × 2 confusion the supplier ratings and the rank of the BWM. In this
matrix for the gradient boosting algorithm. Here we had way get the feature importance score, which shows how
a total of 12 predicted values that aligned the actual value the data are related compared to all the suppliers. Now
for true positive (TP) and true negative (TN). One data in our case, as the rank was used for the training dataset,
point was falsely predicted for the Gradient Boosting the less the score the better the supplier is. To visualize

Algorithm 1. Gradient boosting.


SUSTAINABLE AND RESILIENT INFRASTRUCTURE 11

Table 7. Final weight calculation using BWM.


Level 1 Weight Rank Level 2 Weight Level 3 Weight Final weight Rank
REL 0.2608 2 Quality 0.3917 REL1 0.13819 0.01412 7
REL2 0.20728 0.02118 4
REL3 0.20728 0.02118 5
REL4 0.08291 0.00847 8
REL5 0.33291 0.03401 3
REL6 0.03140 0.00320 9
Accuracy 0.6083 REL7 0.27272 0.04327 2
REL8 0.63636 0.10098 1
REL9 0.09090 0.01442 6
EX 0.1565 4 Delivery 0.5009 FLEX1 0.6287 0.04929 2
FLEX2 0.3713 0.02911 3
Deliver proper quantities 0.4991 FLEX3 1.0 0.07811 1
RES 0.1565 3 Cycle time 0.4125 RES1 0.68 0.04390 2
RES2 0.24 0.01549 4
RES3 0.08 0.00516 5
Quality supplied 0.5875 RES4 0.5388 0.04954 1
RES5 0.4612 0.04241 3
COS 0.0782 5 Transportation cost 0.1 COS1 1.0 0.00783 4
Material cost 0.75 COS2 0.51943 0.03048 1
COS3 0.48056 0.02821 2
Financial status 0.15 COS4 1.0 0.01173 3
DT 0.0260 7 Ability 0.4358 DT1 0.55095 0.00626 2
DT2 0.44904 0.00510 3
Method 0.0769 DT3 0.79259 0.00159 5
DT4 0.14074 0.00028 6
DT5 0.06666 0.00013 7
Systems 0.4872 DT6 0.29606 0.00376 4
DT7 0.70393 0.00894 1
IS 0.0347 6 Integration 0.0666 IS1 0.08333 0.00019 9
IS2 0.17857 0.00041 8
IS3 0.73809 0.00171 6
Content 0.55 IS4 0.5408 0.01034 1
IS5 0.4592 0.00878 2
Documentation 0.3833 IS6 0.04166 0.00055 7
IS7 0.33333 0.00444 4
IS8 0.41666 0.00555 3
IS9 0.20833 0.00277 5
SER 0.2869 1 Customer relationship management 0.3484 SER1 0.6333 0.06331 3
SER2 0.3667 0.03666 4
After-sales service 0.6515 SER3 0.4598 0.08596 2
SER4 0.5402 0.10099 1

the result, we converted the score to an acceptance score normalized the acceptance score within the range of 0
which made it easy for us to understand that the highest to 1, where 1 represents the best supplier. Table 8 shows
weight is the best supplier. Furthermore, we have the supplier’s acceptance score for the gradient boosting

Figure 5. Confusion matrix of gradient boosting algorithm.


12 M. D. M. KHAN ET AL.

Table 8. Acceptance score for gradient boosting.


Supplier (S) Weight Rank Supplier (S) Weight Rank
S20 1.0 1 S4 0.963368 14
S3 1.0 1 S8 0.962932 15
S27 0.991008 2 S2 0.962827 16
S17 0.983849 3 S26 0.962017 17
S13 0.982626 4 S11 0.957659 18
S28 0.979898 5 S6 0.955517 19
S25 0.976302 6 S18 0.955432 20
S23 0.974312 7 S10 0.952034 21
S16 0.973968 8 S1 0.951576 22
S24 0.972216 9 S5 0.94989 23
S14 0.971372 10 S22 0.946033 24
S19 0.969882 11 S15 0.942426 25
S7 0.969348 12 S12 0.928714 26
S9 0.965242 13 S21 0.89955 27

algorithm and the acceptance scores in descending calculation for the criterion of each level of metrics, we
order. applied the gradient boosting ML algorithm to rank the
supplier. Table 8 represents the acceptance scores and
the rank of the 28 suppliers in descending order. From
5. Discussion of the results
Table 8, we can also evaluate the performance of the
The section provides an explanation and interpretation suppliers. Among all the 28 suppliers, Supplier number
of the results by comparing them with the previous 20 (S20) and supplier number 3 (S3) ranked at the
studies. Previously many research works have been con­ position of 1 with an acceptance score of 1.0. So, we
ducted where the SCOR model with the integration of can conclude that S21 and S3 are the best performing
MCDM has been used for supplier selection and very supplier. Apart from that, Supplier number 21 (S21)
few research works have been performed only using ranked 27 with an acceptance score of 0.89955, which
machine learning techniques for supplier selection and represents the weakest performed supplier among all the
performance evaluation. Previously Wilson et al. (2020) 28 suppliers. To measure the performance of our imple­
implemented random forest and Zhao et al. (2021) used mented gradient boosting model, we have calculated the
K fold cross-validation machine learning algorithm for accuracy and error rate of our algorithm from the con­
supplier selection. In this perspective, the proposed fusion matrix. Figure 5 shows the values of true positive
model in this paper is the first unified approach, where = 5, false positive = 1, false negative = 0, and true nega­
the SCOR 4.0 model with the integration of gradient tive = 6. So, our implemented gradient boosting model
boosting machine learning algorithm has been imple­ has an accuracy of 91.6% in supplier selection and the
mented for supplier selection. In this research, we con­ error rate is 8.4%. The performance calculation of gra­
sidered 28 different raw material suppliers of a local dient boosting algorithm of our research has shown in
pharmaceutical company. For this, we grouped the Table A2 in the appendix section. Sepehri (2020) used
main criterion of the suppliers under 7 categories the K fold cross-validation machine learning technique
which are reliability, flexibility, cost, responsiveness, for supplier selection, where the researchers were able to
digital technology, information systems, and services. achieve 92.83% accuracy. Because their dataset was lar­
Extra metrics, such as digital technology and informa­ ger compared to ours. However, in our study, we were
tion systems, are included in the SCOR model to con­ able to achieve the precession percentage of 83.33%
struct the SCOR 4.0 model with a hierarchical metric which measures the correctly predicted number of sup­
framework. The BWM has been used to determine the pliers, whereas the precision percentage achieved by
most and least important criteria of the SCOR 4.0 (Sepehri, 2020) was 70.1%. The findings of the experi­
model. We developed the level 1 metrics based on relia­ ment demonstrate that using SCOR 4.0 in combination
bility, flexibility, cost, responsiveness, digital technol­ with machine learning can assist in identifying risk-free
ogy, information systems, and services. Level 2 and suppliers based on any supplier dataset. The decision-
level 3 metrics have been developed based on level 1 makers in charge of policy can acquire evaluations of
metrics. As discussed earlier, we used BWM to find out which suppliers or combinations of suppliers are the
the weights of level 2 and level 3 metrics. For this most critical in terms of the risks and can impact supply
purpose, we used 9 different linguistic variables to per­ chain performance. Therefore, the implementation of
form the pairwise comparison between the criterion of resilient and sustainable supplier selection in conjunc­
each level of metrics. After the final weightage tion with artificial intelligence-based decision-making
SUSTAINABLE AND RESILIENT INFRASTRUCTURE 13

strategies has the capability of composing a reliable sys­ conclusion, the proposed methodology contributes sig­
tem of supplier selection in any manufacturing nificantly to academic and industrial environments.
environment.
6.2. Implication for company executives and
policymakers
6. Implications
This research demonstrates an in-depth insight into
This section discusses the implications of this research
choosing the most potential suppliers for
from theoretical, practical, and company perspectives.
a pharmaceutical company. This research will help pol­
icymakers in an organization determine which potential
suppliers can be included in their supply chain. This
6.1. Theoretical contribution and practical
study’s methodology will be a useful resource for deci­
implication
sion-makers seeking to comprehend the impact of
This investigation has several implications for supplier diverse supplier selection criteria and sub-criteria. This
development strategies. The primary objective of sup­ snapshot would act as a standard for firms in the same
plier development is to identify prospective suppliers field and can enhance the company’s overall growth and
who can enhance a company’s past performance and future capabilities. The methodologies described in this
future capabilities. Firstly, machine learning can study give a lens for purchasing corporations to not only
improve supply chain risk management methods strengthen existing connections but also to uncover
because it permits the minimization of complexities potentially high-value partnerships in a supplier net­
in risk management models through the analysis of work based on historical profiles and performance out­
historical data and predicted outcomes, which can sub­ comes of any supplier.
stitute a variety of simplifying hypotheses. Secondly,
because the proposed model is based on a process of
7. Conclusion and future works
learning, it has the potential to make the decision-
making process more flexible and can dynamically With the advancement of globalization, it is becoming
analyze previous data to make better decisions. increasingly apparent that the supply chain must incor­
Finally, the methodologies illustrated in this cannot porate resilience to gain a sustainable competitive ben­
only can improve existing connections but also efit. The interconnected imperatives of resilience to
uncover potentially high-value partnerships in disruption and sustainability in evaluating and selecting
a supplier network based on historical patterns of pro­ suppliers’ performance have become a serious strategic
files and value outcomes. Compared to the existing problem. In recent years, environmental pressures have
literature on resilient and sustainable supplier selec­ made the supply chain and its management essential for
tion, this study appears to be the first to use SCOR organizations to remain competitive in the marketplace.
4.0 and machine learning to select the most potential However, a company’s supply chain is exposed to
supplier. This study also demonstrates the effectiveness a greater number of risks, including those posed by
and scalability of machine learning in the domain of nature, humans, and technology. These risks include
supply chain management. The methodology presented pandemics, floods, earthquakes, accidents involving
in this research has academic as well as industrial uses. transportation, labor strikes, and terrorist attacks.
The insights gained from this artificial intelligence- These disasters cause supply chain disruptions, which
based setting show both the flexibility and power of are detrimental to organizations due to lost productiv­
machine learning techniques for making decisions in ity, revenue, competitive advantage, profitability, etc. To
supplier management that are flexible and proactive. protect the buyer from shortages and disruptions, it is
Companies can be more efficient with their resource essential to implement a resilient supply chain strategy
allocation when they have a clear understanding of for supplier selection. The fundamental objective of this
their suppliers’ different profiles and divide them into investigation is to develop a model that incorporates
groups that are easy to manage. Additionally, this a novel method for the selection of resilient and sustain­
methodology can be applied in a variety of industrial able suppliers. The proposed methodology is divided
scenarios that can assist policymaker’s intake decisions into two phases. At the very first, the SCOR 4.0 model
while taking into account the choice of suppliers under was created and the evaluation criteria’s relative weights
diverse circumstances. In addition, humanitarian orga­ were solved by using BWM. The final weights of those
nizations operating mostly in distant places can gain criteria are then employed to rank the best and the most
direct insights from this study’s findings. In important criteria of each of the group metrics. In
14 M. D. M. KHAN ET AL.

the second phase, the gradient boosting ML algorithm is Niamat Ullah Ibne Hossain is an assistant professor in the
implemented to rank the supplier based on the criteria Department of Engineering Management at Arkansas State
of the group metrics. The cognitive ambiguity of experts University, USA. His main research interests include machine
learning, model-based systems engineering (MBSE)/SysML,
can be collected separately and then added to the model, data analytics, systems dynamics simulation, and systems
which is one of the advantages of the hybrid method. resilience, risk & sustainability management.
The SCOR model is more flexible when it comes to
processing information, which is another competitive
advantage of the method shown.
ORCID
In this research, there are some limitations. We have
considered the supplier only related to pharmaceutical Niamat Ullah Ibne Hossain http://orcid.org/0000-0002-
sectors. So, in the future other sectors can be explored. 6775-585X
Apart from that, we have applied only a gradient boost­
ing algorithm for supplier ranking. In the future,
References
researchers can combine the proposed model with
other Machine learning algorithms like Random Forest Abdulla, A., Baryannis, G., & Badi, I. (2019). Weighting the
(RF), Decision Tree (DT), Support Vector Machine key features affecting supplier selection using machine
(SVM), etc. Additionally, the usage of rule-based sys­ learning techniques. International Conference on
Transport and Logistics. University of Niš. December, 1–6.
tems learning algorithms such as deep learning https://doi.org/10.20944/preprints201912.0154.v1
approaches, which could be useful in identifying numer­ Alavi, B., Tavana, M., & Mina, H. (2021). A dynamic decision
ous causalities and enhancing model performance, support system for sustainable supplier selection in circu­
would also be advantageous. lar economy. Sustainable Production and Consumption, 27
(February), 905–920. https://doi.org/10.1016/j.spc.2021.
02.015
Allgurin, A., & Filip, K. (2018). Exploring machine learning
Disclosure statement
for supplier selection-A case study at bufab Sweden AB.
The Coalition for Disaster Resilient Infrastructure (CDRI) 1–81.
reviewed the anonymised abstract of the article, but had no Amindoust, A. (2018). A resilient-sustainable based supplier
role in the peer review process nor the final editorial decision. selection model using a hybrid intelligent method.
Computers & Industrial Engineering, 126(September),
122–135. https://doi.org/10.1016/j.cie.2018.09.031
Amiri, M., Hashemi-Tabatabaei, M., Ghahremanloo, M.,
Funding Keshavarz-Ghorabaee, M., Zavadskas, E. K., &
The Article Publishing Charge (APC) for this article is funded Banaitis, A. (2020). A new fuzzy BWM approach for eval­
by the Coalition for Disaster Resilient Infrastructure (CDRI). uating and selecting a sustainable supplier in supply chain
management. International Journal of Sustainable
Development and World Ecology, 00(00), 1–18. https://doi.
org/10.1080/13504509.2020.1793424
Notes on contributors Ayyildiz, E., & Taskin Gumus, A. (2021). Interval-valued
pythagorean fuzzy AHP method-based supply chain per­
Md Muzahid Khan is a Lecturer in the Department of
formance evaluation by a new extension of SCOR model:
Military Institute of Science and Technology, Dhaka,
SCOR 4.0. Complex & Intelligent Systems, 7(1), 559–576.
Bangladesh. His research interests include supply chain man­
https://doi.org/10.1007/s40747-020-00221-9
agement, sustainability quantification, healthcare and
Badi, I., & Ballem, M. (2018). Supplier selection using rough
machine learning.
bwm-mairca model : A case study in pharmaceutical sup­
Imranul Bashar completed his bachelor's in Industrial and plying in libya. Decision Making: Applications in
Production Engineering from the Military Institute of Science Management and Engineering, 1(2), 1–18. https://doi.org/
and Technology, Dhaka, Bangladesh. His research interests 10.31181/dmame1802016b
include supply chain management, machine learning, and Batta, M. (2020). Machine learning algorithms - a review.
healthcare. International Journal of Science and Research (IJ interna­
tional Journal of Science and Research (IJ, 9(1), 381. https://
Golam Morshed Minhaj completed his bachelor's in
doi.org/10.21275/ART20203995
Industrial and Production Engineering from the Military
Bentéjac, C., Csörgő, A., & Martínez-Muñoz, G. (2021).
Institute of Science and Technology, Dhaka, Bangladesh. His
A comparative analysis of gradient boosting algorithms.
research interests include supply chain analytics and
In Artificial Intelligence Review (Vol. 54, Issue 3). Springer
healthcare.
Netherlands. https://doi.org/10.1007/s10462-020-09896-5
Absar Ishraq Wasi completed his bachelor's in Industrial and Cavalcante, I. M., Frazzon, E. M., Forcellini, F. A., &
Production Engineering from the Military Institute of Science Ivanov, D. (2019). A supervised machine learning approach
and Technology, Dhaka, Bangladesh. His research interests to data-driven simulation of resilient supplier selection in
include supplier management, data science, and optimization. digital manufacturing. International Journal of Information
SUSTAINABLE AND RESILIENT INFRASTRUCTURE 15

Management, 49, 86–97. https://doi.org/10.1016/j.ijin A novel BWM–GAIA framework. Annals of Operations


fomgt.2019.03.004 Research. https://doi.org/10.1007/s10479-022-04710-7
Chamola, V., Hassija, V., Gupta, V., & Guizani, M. (2020). Islam, S., Amin, S. H., & Wardley, L. J. (2021). Machine
A comprehensive review of the COVID-19 pandemic and learning and optimization models for supplier selection
the role of IoT, drones, AI, blockchain, and 5G in managing and order allocation planning. International Journal of
its impact. IEEE Access, 8, 90225–90265. https://doi.org/10. Production Economics, 242, 108315. https://doi.org/10.
1109/ACCESS.2020.2992341 1016/J.IJPE.2021.108315
Craighead, C. W., Ketchen, D. J., & Darby, J. L. (2020). Jain, V., Kumar, S., Mostofi, A., & Arab Momeni, M. (2022).
Pandemics and supply chain management research: Sustainability performance evaluation of the E-waste
Toward a theoretical toolbox*. Decision Sciences, 51(4), closed-loop supply chain with the SCOR model. Waste
838–866. https://doi.org/10.1111/DECI.12468 Management, 147, 36–47. https://doi.org/10.1016/J.
Delipinar, G. E., & Kocaoglu, B. (2016). Using SCOR model to WASMAN.2022.05.010
gain competitive advantage: A literature review. Procedia - Judith, H., & Daniel, K. (2018). Machine learning for
Social and Behavioral Sciences, 229, 398–406. https://doi. dummies. Journal of the American Society for Information
org/10.1016/j.sbspro.2016.07.150 Science, 35(5). https://doi.org/10.1002/asi.4630350509
Didier, M., Broccardo, M., Esposito, S., & Stojadinovic, B. Lima-Junior, F. R., & Carpinetti, L. C. R. (2016). Combining
(2018). A compositional demand/supply framework to SCOR® model and fuzzy TOPSIS for supplier evaluation
quantify the resilience of civil infrastructure systems and management. International Journal of Production
(Re-CoDes). Sustainable and Resilient Infrastructure, 3(2), Economics, 174, 128–141. https://doi.org/10.1016/j.ijpe.
86–102. https://doi.org/10.1080/23789689.2017.1364560 2016.01.023
Dorogush, A. V., Ershov, V., & Gulin, A. (2018). CatBoost: Lu, Q., Goh, M., & De Souza, R. (2016). A SCOR framework to
Gradient boosting with categorical features support. https:// measure logistics performance of humanitarian
doi.org/10.48550/arXiv.1810.11363 organizations. Journal of Humanitarian Logistics and
Es-Satty, A., Lemghari, R., & Okar, C. (2020). Supply chain Supply Chain Management, 6(2), 222–239. https://doi.org/
digitalization overview SCOR model implication. 2020 13th 10.1108/JHLSCM-09-2015-0038
International Colloquium of Logistics and Supply Chain Luitzen, D. B., Eva, L., & Pierangela, M. (2001). A review of
Management, LOGISTIQUA 2020, 2–4. https://doi.org/10. methods supporting supplier selection. European Journal of
1109/LOGISTIQUA49782.2020.9353936 Purchasing & Supply Management, 80(3), 66–73. https://
Ghadimi, P., Wang, C., Lim, M. K., & Heavey, C. (2019). doi.org/10.1002/ad.1077
Intelligent sustainable supplier selection using multi-agent Mañay, L. O. R., Guaita-Pradas, I., & Marques-Perez, I.
technology: Theory and application for industry 4.0 supply (2022). Measuring the supply chain performance of the
chains. Computers & Industrial Engineering, 127 floricultural sector using the SCOR model and
(September 2018), 588–600. https://doi.org/10.1016/j.cie. a multicriteria decision-making method. Horticulturae, 8
2018.10.050 (2), 168. https://doi.org/10.3390/horticulturae8020168
Ghamari, R., Mahdavi-Mazdeh, M., & Ghannadpour, S. F. Mirkouei, A., & Haapala, K. R. (2014). Integration of machine
(2021). Resilient and sustainable supplier selection via learning and mathematical programming methods into the
a new framework: A case study from the steel industry. In biomass feedstock supplier selection process. FAIM 2014 -
Environment, development and sustainability (Vol. 24, Proceedings of the 24th International Conference on Flexible
Issue 8). Springer Netherlands. https://doi.org/10.1007/ Automation and Intelligent Manufacturing: Capturing
s10668-021-01872-5 Competitive Advantage via Advanced Manufacturing and
He, J., & Zhu, J. (2022). Evaluation on risk factors of elderly Enterprise Transformation, January, 443–450. https://doi.
services from the perspective of integrated SCOR model. org/10.14809/faim.2014.0443
International Journal of Information Systems in the Service Munson, A. D., & Santa-Eulalia, L. A. D. (2014). A Systematic
Sector, 14(1), 1–18. https://doi.org/10.4018/IJISSS. literature review of the supply chain operations reference
2022010102 (SCOR) model application with special attention to envrionn­
Hossain, N. U. I., Nur, F., & Jaradat, R. M. (2016). An analy­ mental issues a systematic literature review of the supply
tical study of hazards and risks in the shipbuilding industry. chain operations reference (SCOR) model application with
In Proceedings of the international annual conference of the spec. January. https://doi.org/10.13140/RG.2.1.3618.7606
American society for engineering management (pp. 1–8). Mustafa, S. K., Engin, E., Belkız, T., & Turan, P. (2020).
American Society for Engineering Management. Supplier selection with machine learning algorithms.
Huang, T., Chen, Z., Wang, S., & Jiang, D. (2021). Efficiency Logistics 4 0, (Issue December 2020). https://doi.org/10.
evaluation of key ports along the 21st-century maritime silk 1201/9780429327636-3
road based on the DEA–SCOR model. Maritime Policy & Nazim, R., & Yaacob, R. A. I. R. (2017). Criteria for supplier
Management, 48(3), 378–390. https://doi.org/10.1080/ selection: An application of AHP-SCOR integrated model
03088839.2020.1773558 (ASIM). International Journal of Supply Chain
Hwang, Y. D., Lin, Y. C., & Lyu, J. (2008). The performance Management, 6(3), 284–290.
evaluation of SCOR sourcing process—the case study of Oroojeni Mohammad Javad, M., Darvishi, M., & Oroojeni
Taiwan’s TFT-LCD industry. International Journal of Mohammad Javad, A. (2020). Green supplier selection for
Production Economics, 115(2), 411–423. https://doi.org/10. the steel industry using BWM and fuzzy TOPSIS: A case
1016/J.IJPE.2007.09.014 study of Khouzestan steel company. Sustainable Futures, 2
Ishizaka, A., Khan, S. A., Kheybari, S., & Zaman, S. I. (2022). (October 2019), 100012. https://doi.org/10.1016/j.sftr.2020.
Supplier selection in closed loop pharma supply chain: 100012
16 M. D. M. KHAN ET AL.

Pamučar, D., Ecer, F., Cirovic, G., & Arlasheedi, M. A. (2020). Expert Systems with Applications, 61, 129–144. https://doi.
Application of improved best worst method (BWM) in org/10.1016/J.ESWA.2016.05.027
real-world problems. Mathematics, 8(8), 1342. https://doi. Wendler-Bosco, V., & Nicholson, C. (2020). Port disruption
org/10.3390/MATH8081342 impact on the maritime supply chain: A literature review.
Park, J., Shin, K., Chang, T. W., & Park, J. (2010). An inte­ Sustainable and Resilient Infrastructure, 5(6), 378–394.
grative framework for supplier relationship management. https://doi.org/10.1080/23789689.2019.1600961
Industrial Management & Data Systems, 110(4), 495–515. Wilson, V. H., Prasad, A. N. S., Shankharan, A., Kapoor, S., &
https://doi.org/10.1108/02635571011038990 Rajan, J. A. (2020). Ranking of supplier performance using
Pilkington, S. F., & Mahmoud, H. N. (2016). Using artificial machine learning algorithm of random forest. International
neural networks to forecast economic impact of Journal of Advanced Research in Engineering and
multi-hazard hurricane-based events. Sustainable and Technology, 11(5), 298–308. https://doi.org/10.34218/
Resilient Infrastructure, 1(1–2), 63–83. https://doi.org/10. IJARET.11.5.2020.031
1080/23789689.2016.1179529 Wu, C., & Barnes, D. (2018). Design of agile supply chains
Raut, R. D., Bhasin, H. V., & Kamble, S. S. (2012). Supplier including the trade-off between number of partners and
selection using integrated multi-criteria decision-making reliability. International Journal of Advanced
methodology. International Journal of Operational Manufacturing Technology, 97(9–12), 3683–3700. https://
Research, 13(4), 359–394. https://doi.org/10.1504/IJOR. doi.org/10.1007/S00170-018-2205-5
2012.046223 Zangoueinezhad, A., Azar, A., & Kazazi, A. (2011). Using
Rezaei, J. (2015). Best-worst multi-criteria decision-making SCOR model with fuzzy MCDM approach to assess com­
method. Omega, 53, 49–57. https://doi.org/10.1016/J. petitiveness positioning of supply chains: Focus on ship­
OMEGA.2014.11.009 building supply chains. Maritime Policy & Management, 38
Roshan, V., Stewart, J. H. M., Joseph, R., & Stewart, H. M. (1), 93–109. https://doi.org/10.1080/03088839.2010.533715
(2022). Optimal ratio for data splitting. Stat Anal Data Min: Zeydan, M., Çolpan, C., & Çobanoģlu, C. (2011). A combined
The ASA Data Sci Journal, 15(4), 15. https://doi.org/10. methodology for supplier selection and performance
1002/sam.11583 evaluation. Expert Systems with Applications, 38(3),
Sepehri, S. (2020). Supplier selection and relationship man­ 2741–2751. https://doi.org/10.1016/j.eswa.2010.08.064
agement : An application of machine learning techniques. Zhang, X. -D. (2017). Chapter 6 machine learning. A Matrix
Ontario, Canada: Brock University. https://dr.library. Algebra Approach to Artificial Intelligence, 45(13), 223–235.
brocku.ca/handle/10464/14831 Zhang, R., Li, J., Wu, S., Meng, D., & Deng, Y. (2016).
Spieske, A., Gebhardt, M., Kopyto, M., & Birkel, H. (2022). Learning to select supplier portfolios for service supply
Improving resilience of the healthcare supply chain in chain. Plos One, 11(5), e0155672. https://doi.org/10.1371/
a pandemic: Evidence from Europe during the COVID-19 journal.pone.0155672
crisis. Journal of Purchasing and Supply Management, 28 Zhaoa, P., Jia, S., & Xuea, Y. (2021, October). Resilient-
(5), 100748. https://doi.org/10.1016/J.PURSUP.2022. sustainable supplier selection and order allocation under
100748 multi-tier supplier network. Researchgate.Net, https://www.
Stević, Ž., Pamučar, D., Puška, A., & Chatterjee, P. (2020). researchgate.net/profile/Shoufeng-Ji/publication/
Sustainable supplier selection in healthcare industries using 355788632_Resilient-sustainable_supplier_selection_and_
a new MCDM method: Measurement of alternatives and order_allocation_under_multi-tier_supplier_network
ranking according to COmpromise solution (MARCOS). /links/617e09c4eef53e51e10c4b17/Resilient-sustainable-
Computers & Industrial Engineering, 140, 106231. https:// supplier-selection-and-order-alloca
doi.org/10.1016/j.cie.2019.106231 Zhao, L., Qi, W., Zhu, M., & Xie, L. (2021). A study of supplier
Tavana, M., Fallahpour, A., DiCaprio, D., & Santos-Arteaga, selection method based on SVM for weighting expert
F. J. (2016). A hybrid intelligent fuzzy predictive model evaluation. Discrete Dynamics in Nature and Society, 2021,
with simulation for supplier evaluation and selection. 1–11. https://doi.org/10.1155/2021/8056209
SUSTAINABLE AND RESILIENT INFRASTRUCTURE 17

Appendix

Table A1. Glossary of level 2 and 3 metrics.


Abbreviation Elaboration Abbreviation Elaboration
REL1 Quality certification DT2 Collaboration
REL2 Product specification DT3 Digital marketing
REL3 Defect percentage DT4 Machine learning
REL4 Packing and labeling DT5 Other methods
REL5 Transportation quality DT6 Facility
REL6 Research, Development, and Innovation DT7 Consistency
REL7 Order fulfillment IS1 Integrated database
REL8 Delivery performance IS2 Integrated interface
REL9 Document accuracy IS3 Partner interface
FLEX1 On-time delivery IS4 Consistent data
FLEX2 Delivery performance IS5 Tactical Level
FLEX3 Document accuracy IS6 Supplier information
RES1 Delivery cycle time IS7 Delivery documentation
RES2 Source cycle time IS8 Qualification and validation
RES3 Delivery retail cycle time IS9 Accord certification
RES4 Supply time SER1 Communication process
RES5 Lead time SER2 Tracking management system
COS1 Shipping SER3 Warranty
COS2 Mode of transaction SER4 Return Policy
COS3 Price DT1 Dynamism
COS4 Previous record DT2 Collaboration
DT1 Dynamism

Table A2. Performance calculation of gradient boosting algorithm.


Performance properties Equations Percentage
Accuracy True positiveþTrue negative 91.6%
True falseþTrue negative True positiveþTrue negative
Error 1-Accuracy 8.4%
Precession True positive 83.33%
True positiveþFalse positive

You might also like