Page 179

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

Decision Support System for Zakat Asnaf Selection among Uitm

Melaka Students Using Artificial Neural Networks

Ismadi Md Badarudin

Ameera Iman Hassan.,

Suzana Ahmad.,

Yuzi Mahmud.,

Khairunnisa Abd

Samad.,

Noor Afni Deraman

Fakulti Sains Komputer dan Matematik UiTM

Cloud Mile Sdn. Bhd,

Fakulti Pengurusan dan Perniagaan UiTM

*Corresponding author

DOI: https://dx.doi.org/10.47772/IJRISS.2025.923MIC3ST250015

Received: 12 August 2025; Accepted: 20 August 2025; Published: 24 October 2025

ABSTRACT

This study developed a Decision Support System for Zakat Asnaf Selection (DSSZAS) to address

inefficiencies in the manual distribution of zakat among students at Universiti Teknologi MARA (UiTM)

Cawangan Melaka. The current process faces challenges in accurately identifying eligible asnaf and

distributing promptly. Therefore, to solve this, the DSSZAS leverages Artificial Neural Networks (ANN) to

automate the classification of students into asnaf categories (faqr, miskin, and fisabilillah) based on

socioeconomic data. The system was designed and trained with historical data using the Waterfall

methodology. A comparison method was deployed between the generated result and human decision to test the

result reliability. It achieves an accuracy rate of 1.0% with a minimized Mean Squared Error (MSE) of 0.06.

The system significantly reduces human bias and enhances efficiency through automated decision-making and

email notifications that inform students of their application status., DSSZAS strengthens the fairness and

reliability of zakat distribution by providing a transparent and data-driven approach.

Keywords: Artificial Neural Networks (ANN), Classification, Decision Support System (DSS), Zakat

Distribution

INTRODUCTION

Zakat, the third pillar of Islam, is an obligatory charitable contribution that Muslim must give annually from

their wealth and income. It plays a critical role in the welfare of the Muslim community, aiming to reduce

poverty and inequality. By distributing wealth to the less fortunate, zakat fosters social harmony and financial

stability (Othman, S. H. B. et. al, (2020)). Each Muslim individual’s possessions, include wealth and income,

should be considered to the obligatory annual payment under Islamic law (Saad, A. Y. Q., & Al Foori, A. M.,

2020). There is a zakat institution in the Muslim community engage with collects and distribute the zakat to

accomplish establishing the welfare of people (Othman et al., 2020). While those people that are involved in

receiving zakat aid during distribution activities are called zakat recipients or asnaf, which has 8 categories.

Zakat funds is redistribute to eight groups from eligible asnaf which are poor, needy, amil, muallaf, al-

Gharimin, fisabilillah, and Ibnussabil notably the utmost focus are the poor and needy (Zulkifli et al., (2021)).

In Malaysia, the management of zakat is carried out by each State Islamic Religious Council, with zakat

institutions responsible for collecting and distributing funds. These institutions follow a structured process to

ensure that zakat reaches the rightful recipients, known as asnaf, who fall into eight categories including the

faqr (poor), miskin (needy), and fisabilillah (in the cause of Allah) and others deserving of assistance (Radzi,

N. M., & Kenayathulla, H. B. (2017)). While efforts have improved zakat collection in several states, the

Page 180

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

distribution process remains complex, often requiring manual assessments and interviews to determine

eligibility, which can delay assistance to those in need (Salleh, M. C. M., & Chowdhury, M. A. M. (2020)).

At Universiti Teknologi MARA (UiTMCM), a similar system exists for distributing zakat to students.

UiTMCM, one of the university’s branches, manages zakat distribution through the Wakaq, Zakat, and Infaq

(EL-WAZIF) Unit. Here, students from lower-income households (B40) who meet the criteria for asnaf—such

as faqr, miskin, and fisabilillah—can apply for zakat assistance. However, the current process involves manual

interviews and decision-making, which presents challenges in terms of efficiency and timely distribution.

This paper addresses the challenges of zakat distribution at UiTM Cawangan Melaka by developing a

decision-making model using Artificial Neural Networks (ANN). It highlights the inefficiencies of the manual

process for selecting eligible asnaf students, which is time-consuming and complex. The proposed Decision

Support System for Zakat Asnaf Selection (DSSZAS) uses ANN to automate student classification and

prioritisation based on socioeconomic data to solve this. The ANN model is trained with historical data,

providing a more accurate and fair distribution of zakat. The system improves the ranking of applicants,

reduces processing time, and enhances transparency.

LITERATURE REVIEW

Zakat Implementation in Malaysian Universities

Several universities in Malaysia have developed tailored systems to manage zakat distribution effectively for

their students and staff. Universiti Putra Malaysia (UPM), through its Centre for Management of Waqf, Zakat,

and Endowment (WAZAN), manages zakat in collaboration with the Selangor Zakat Board. UPM is

authorised to distribute zakat to five categories of asnaf—fakir (destitute), miskin (poor), muallaf (new

converts), ibn sabil (wayfarers), and fisabilillah (in the cause of Allah). The UPM-ID system facilitates this

process by allowing students to submit applications online, which are validated through interviews before

approval. The process ensures transparency and accountability, with notifications sent via email before the

zakat aid is transferred to student bank accounts (Pusat Pengurusan Wakaf, Zakat dan Infak, 2022).

Similarly, Universiti Teknologi Malaysia (UTM) and Universiti Tun Hussein Onn Malaysia (UTHM) have

adopted digital systems to manage student zakat distribution. UTM’s Islamic Center administers the Zakat

Distribution System through the MyUTM Student portal, where students can apply for various forms of Zakat

aid, including tuition and subsistence assistance, with results communicated via email (Hamzah, N. B. et. al.

(2021)). UTHM, on the other hand, uses student portal for zakat application submission including show the

results. Any notification is shared through social media such as Whatsapp and Telegram (Pusat Islam Unit

Pengurasan Zakat UTHM (2025). Although these systems differ in their features—such as notification

methods and had kifayah calculations, they all aim to streamline zakat distribution and provide timely financial

assistance to eligible students.

Zakat Implementation in UiTM Melaka

UiTM Cawangan Melaka has also introduced a zakat distribution system to support its student body,

particularly those from the B40 income group, who face financial challenges. Managed by the Islamic Affairs

Unit (UHEI) and supported by the Academy of Contemporary Islamic Studies (ACIS), the zakat distribution

process is overseen by three Islamic Affairs Officers across three campuses: Kampus Alor Gajah, Kampus

Bandaraya Melaka, and Kampus Jasin. The zakat funds are allocated to students who fall under faqr, miskin,

fisabilillah, and muallaf. Eligible students receive zakat funds every semester, helping alleviate their financial

burdens.

Initially, zakat distribution was done manually, but starting in the 2021/2022 session, UiTM Melaka

introduced the eZakat system within the eHEP platform to digitize the process. Students complete an eight-

section application covering personal, family, and financial details, with the system automatically calculating

the kifayah limit based on family income and dependents. Interviews are conducted with students whose

applications contain uncertainties, and final decisions are made following an approval process involving

Page 181

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

campus management. Once approved, zakat funds are transferred to the student’s bank accounts, typically

during the final examination period. The shift to eZakat has improved efficiency and transparency, enabling

quicker processing and more straightforward communication with students via the UHEI Facebook page and

the eZakat system.

Artificial Neural Networks (ANN) in Zakat Implementation

Artificial Neural Networks (ANN), a powerful Artificial Intelligence (AI) tool, offer a transformative approach

to enhancing zakat implementation, particularly in automating the identification and prioritization of asnaf.

ANN models are adept at analyzing complex socioeconomic data, such as income levels and household sizes,

to classify recipients more precisely. By learning from historical data, ANN systems can continuously refine

their decision-making process, ensuring that zakat funds are distributed based on objective, data-driven

criteria. This minimizes human bias, speeds up decision-making, and ensures that resources are allocated

equitably, with the most vulnerable recipients receiving priority (Bahri E. S. et. al, (2022))

Moreover, ANN improves the overall efficiency of zakat management by streamlining operations and reducing

fraud. These systems can detect anomalies in application data, identify fraudulent claims and ensure that funds

are directed to those genuinely in need (Alhajeri R. and Alhashem A. (2023)). The scalability of ANN makes it

ideal for large-scale zakat systems, processing thousands of applications in a fraction of the time required for

manual assessments. When combined with blockchain technology, ANN further enhances transparency by

creating tamper-proof records of zakat transactions, enabling stakeholders to track funds with greater

accountability. Integrating ANN into zakat distribution systems makes the entire process more efficient, fair,

and secure, significantly improving the system's ability to alleviate poverty and support vulnerable

communities (Miftahur A. H. et. al. (2020), Sulaiman H. and Jamil N. (2014))

There are two techniques of data obtain that can be manage by the ANN. The unlabelled data is group by

according to similarities among the example inputs and labelled dataset is being classify to train these datas

(Khalil et al., 2019). To distinguish it is from the type of data that are labeled or unlabeled data. The

discovered from numerous studies is classification of neural network is more efficient and more accurate to

tackle real-world problems due to their capability of self-learning and self-adapting (Al-Mubayyed et al.,

2019). Classification phase starts of with the network is trained on a set of paired input and output data to

determine its mapping. After the weight of the connections between neurons are fixed, the network can classify

the new set of data. For unlabelled sample or unordered data, clustering method is used by grouping objects

according to measured intrinsic characteristics or similarity into homogeneous group (Raptodimos & Lazakis,

2018).

RESEARCH METHODOLOGY

This section outlines the approach taken to developing an efficient and transparent zakat distribution system,

leveraging system design principles and Artificial Neural Networks (ANN) to optimize decision-making. The

methodology is divided into two subsections: the first focuses on the design of the Decision Support System

for Asnaf Zakat Selection (DSSZAS), detailing the workflow, data collection, and implementation processes;

the second subsection discusses the use of ANN in enhancing the system’s ability to classify and prioritize

asnaf applicants, ensuring a fair and data-driven distribution of zakat funds.

System Design

The development of the DSSZAS begins with gathering input and analyzing the requirements of the existing

business process. This phase is crucial for understanding the current workflow, which involves determining

eligible asnaf applicants, conducting interviews, and assigning appropriate zakat amounts. This process clearly

defines the system's requirements, forming the foundation for creating relevant diagrams such as the Use Case

Diagram, Data Flow Diagram (DFD), and Domain Class Diagram (DCD).

Once the requirements are collected, the system design phase begins. This involves constructing a Use Case

Diagram to represent the interactions between users such as students, evaluators, and administrators and the

Page 182

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

system. The Data Flow Diagram (DFD) defines the system's boundaries. It outlines how data flows between

external entities and the system as shown in Figure 1 and Figure 2. Additionally, a Domain Class Diagram

(DCD) is created to design the database structure, ensuring that all necessary data elements, such as student

information, evaluation results, and zakat records, are captured efficiently. These diagrams provide a

comprehensive view of how the system will function and how various users will interact with it.

Figure 1: Use case diagram

Figure 2: Context Diagram for Zakat Distribution System

The system flowchart Figure 3, outlines the complete process from application submission to final approval.

Initially, students submit their applications through the eZakat platform, where their eligibility is automatically

filtered based on had kifayah, ensuring that only students from the B40 income group (family income below

RM4,000) proceed. Following this, the applications undergo an evaluation phase where the evaluators review

and resolve any uncertainties in the data. The flowchart demonstrates how Artificial Neural Networks (ANN)

will be used in the final decision-making process. By combining student data with evaluator input, the system

ensures that zakat funds are distributed fairly and accurately. Once the evaluation is completed, the results are

saved in a report format accessible to both administrators and evaluators for final approval. This flowchart

provides a clear, step-by-step depiction of the system's workflow, ensuring an efficient and transparent zakat

distribution process.

Page 183

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

Figure 3: System Flowchart

Artificial Neural Network (ANN) Strategy

The Artificial Neural Network (ANN) strategy is vital in automating and optimizing the zakat allocation

process. It combines student information with evaluator inputs to determine the appropriate zakat distribution.

The ANN model is designed to ensure fairness, accuracy, and efficiency in decision-making, considering

various socioeconomic factors related to students’ financial needs.

The ANN model begins by gathering inputs, including student data such as CGPA, number of siblings,

outstanding loans, and an evaluator-assigned ranking. These inputs are critical to determining the student’s

eligibility as asnaf faqr, miskin, or fisabilillah. Collected data is transformed into a numerical range of 0.0 and

1.0. For example, the asnaf numerical for faqr as 0.9, miskin as 0.5 and fisabilillah as 0.2. As shown in Table

1, the weight for each input is decided by domain experts, with the flexibility to adjust the values as needed.

Domain experts initially assigned input weights, with 0 being the least important and 1 the most, as detailed in

Table 1 (Importance A). However, Importance A yielded poor results during the Artificial Neural Network

(ANN) implementation. Subsequent adjustments revealed that Importance B produced superior results.

Importance B shown that the number of siblings and loan balances hold a higher weight (0.9) compared to

CGPA (0.5), while ranking holds the maximum weight (1). These weights ensure that the most financially

vulnerable students are prioritised during the zakat allocation process.

Table 1: Weight of the Input

Input

Importance A (First Test)

Importance B (Second Test)

CGPA

0.4

0.5

Number of siblings that is studying

0.7

0.9

Reminder of loan

0.8

0.9

Ranking

0.5

*During the first test, the results are poor. As the value changes, this is the most importance that showing less

MSE

Next, the ANN strategy proceeds with data normalization, where input values from the database are

transformed into smaller, standardized values. This step is necessary to improve the efficiency of the neural

network by ensuring that all inputs fall within a comparable range. The data is split into two sets: 80% is used

Page 184

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

for training, while the remaining 20% is reserved for testing the model’s accuracy. Using a sigmoid function,

the feedforward method, is applied to process the data. The resulting output is between 0 and 1, representing

whether the student qualifies as asnaf. A loss function is employed within the model to evaluate errors between

the predicted and actual outputs. To reduce this error, backpropagation is used to adjust the weights within the

neurons, ensuring that the system improves with each iteration. The Mean Squared Error (MSE) function is

employed to quantify the difference between predicted and target values, providing the model’s continuous

learning.

The ANN model is structured with three layers: one input layer, three hidden layers, and one output layer. The

input neurons receive the student data, and the hidden layers process the information, calculating intermediate

outputs. The weights assigned to each neuron are based on the significance of the input, as outlined by the

experts. During this process, the ANN model runs in multiple iterations, adjusting the weights to minimise

errors between the predicted and actual results. As backpropagation update the weights, the model becomes

more accurate in predicting the final classification of asnaf.

The ANN model ultimately outputs a determined classification whether the student falls into one of the three

asnaf categories. After training and testing the model, the system should reliably identify eligible students and

determine the appropriate zakat distribution. This approach ensures that the zakat allocation is based on a data-

driven and unbiased process, improving the fairness and efficiency of the overall system.

RESULT AND DISCUSSION

The developed zakat distribution system DSSZAS is built on the Model-View-Controller (MVC) framework,

which separates the system into three distinct layers: the view (user interface), the controller (business logic),

and the model (data access). This architecture ensures efficient communication between the components,

where the front end interacts with users and sends operations to the controller, while the model handles data

storage, retrieval, and manipulation. The core objective of this system is to build a decision support system that

can accurately classify students into one of three asnaf categories. The system achieves this through integrating

data input from students, staff rankings, and an Artificial Neural Network (ANN) model. Additionally, the

system includes an email notification feature to inform students of the successful completion of their

applications.

The ANN model used in the system was trained using a batch-size method to handle student data. The model

required training data that consisted of previous student information with corresponding asnaf classifications,

as depicted in Figure 4 and Figure 5. During the training phase, the feed-forward method was applied to

process the input data, with the Mean Squared Error (MSE) used to calculate the loss between the predicted

and actual outputs. Backpropagation was employed to minimize this loss by adjusting the neuron weights.

After training, the model was tested to assess its accuracy in classifying students. The initial test with 14 data

points yielded an accuracy of 0.35 with a 4-1-1 architecture. As the training data set increased to 200 data

points, the accuracy improved to 0.39%, and the MSE was reduced to 0.11.

Figure 4: Train Dataset

Page 185

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

Figure 5: Test Dataset

The first training with the initial weight results in an accuracy drop. Perform more testing by changing the

importance of weight to lower importance especially the ranking weight because the data is less varied with

only three distinct values. The ranking weight becomes 0.5 because most of the training data points are

correctly assigned, leading to 1.0 accuracy. To compare the system-generated asnaf classifications with the

panel’s original results, the testing data was modified to incorrect asnaf classification only. During panel

justification, panel can modify the students' ranking to correct the assigned asnaf. After the testing data is

completed, the asnaf classification is compared with the previous results and it comes out similar as shown in

Table 2.

Table 2: Asnaf Test Result

Current Asnaf

Expected Asnaf Result

Test 1

Test 2

Test 3

Dataset 1

Miskin

Faqr

Not Match

Match

Dataset 2

Fisabilillah

Miskin

Not Match

Match

Further refinement of the ANN model included increasing the neurons in the hidden layer. As demonstrated by

the third test with an expanded hidden layer, the accuracy rose to 1.0%, and the MSE was further minimized to

0.06. This indicates the system's performance improves as more data is trained, and the network architecture is

optimized. The final result shows that the ANN model is highly effective in classifying asnaf categories with

an acceptable level of accuracy, as shown in Figure 6. The decision support system generated by the DSSZAS

successfully mirrors the decisions typically made by human evaluators, demonstrating that the system can be

relied upon to automate the selection process while maintaining the integrity and fairness of zakat distribution.

Figure 6: Asnaf Zakat Selection Generated by DSSZAS

Overall, the results confirm that the ANN-based decision support system is a viable solution for automating the

classification and selection of asnaf recipients. The system's accuracy and ability to minimize errors through

backpropagation and MSE demonstrate that it can significantly streamline zakat distribution processes,

Page 186

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

improving efficiency and transparency. By integrating data-driven techniques with automated decision-

making, the system provides a robust and scalable solution for zakat management in educational institutions.

CONCLUSION

This study successfully developed an ANN-based decision support system to automate and optimize zakat

distribution for students in UiTM Cawangan Melaka. By classifying students into asnaf categories of faqr,

miskin and fisabilillah based on socioeconomic data, the system effectively addressed the inefficiencies of

manual zakat processing. The ANN model demonstrated improved decision-making accuracy, reaching up to

1.0% with a minimized Mean Squared Error (MSE) of 0.06, indicating the system’s ability to make reliable

and data-driven decisions. Integrating a Model-View-Controller (MVC) framework ensured smooth interaction

between users, evaluators, and the system’s database, Overall, the system significantly improved the fairness,

efficiency, and transparency of zakat allocation.

Future research could focus on expanding the dataset used for training the ANN model, as larger datasets

might further increase the system’s accuracy and robustness. Integrating real-time data updates and

dynamically adjusting weights could also refine the decision-making process. Additionally, exploring the use

of blockchain technology to enhance security and traceability of transactions could add another layer of

transparency and accountability. Expanding the system to include multiple universities or national zakat

institutions could create a centralized platform for zakat management, benefiting a broader population. These

improvements would continue to enhance the efficiency and equity of zakat distribution, further supporting

social welfare.

ACKNOWLEDGMENTS

This paper is the starting point of Final Year Project conducted at the Faculty of Computer and Mathematic

Sciences, Universiti Teknologi MARA.

REFERENCES

1. Othman, S. H. B., Ismon, A. B. M., & Fahurruzi, M. F. (2020). The Study on The Development of

Asnaf fisabilillah by The Islamic Religious Council of Malacca. International Journal of Academic

Research in Business and Social Sciences, 10(2). https://doi.org/10.6007/ijarbss/v10-i2/6929

2. Radzi, N. M., & Kenayathulla, H. B. (2017). Educational Equity for Malaysian Urban Poor Muslim

Students. Malaysian Online Journal of Educational Management, 5(4), 47–62.

https://doi.org/10.22452/mojem.vol5no4.4

3. Salleh, M. C. M., & Chowdhury, M. A. M. (2020). Technological Transformation in Malaysian Zakat

Institutions. International Journal of Zakat, 5(3), 44–56. https://doi.org/10.37706/ijaz.v5i3.263

4. Pusat Pengurusan Wakaf, Zakat dan Infak (2022, Oct 14). https://wazan.upm.edu.my/permohonan

zakat/permohonan_zakat_pelajar/bagaimana_proses_permohonan_zakat-63697

5. Hamzah, N. B., Masroom, Mohd. N. B., Ruskam, A. B., & Saim, A. I. B. (2021). Penggunaan Wang

Bantuan Zakat Dalam Kalangan Asnaf Fisabilillah Dan Asnaf Gharim Di Universiti Teknologi

Malaysia. AZKA International Journal of Zakat & Social Finance, 1(2), 187–209.

https://doi.org/10.51377/azjaf.vol1no2.37

6. Pusat Islam Unit Pengurasan Zakat UTHM (2025, Mac 21). https://pi.uthm.edu.my/index.php/zakat

7. Bahri E. S., Salamun A, and Arif Z. (2022), Measuring the Effectiveness of Zakat Disbursement at

Amil Zakat Institution Dewan Da’wah Utilizing Zakat Core Principle Approach, Al- Maal: Journal of

Islamic Economics and Banking, vol. 4, no. 1, p. 119, https://doi.org/10.31000/almaal.v4i1.5937.

8. Alhajeri R. and Alhashem A. (2023), Using Artificial Intelligence to Combat Money Laundering,

Intelligent Information Management, vol. 15, no. 4, pp. 284–305,

https://doi.org/10.4236/iim.2023.154014.

9. Miftahur A. H., Leni S, and Kholidah (2020), Ziswaf Collection Strategy on LAZISMU X, Indonesia,

Proceedings of the 1st Borobudur International Symposium on Humanities, Economics and Social

Sciences (BIS-HESS 2019), https://doi.org/10.2991/assehr.k.200529.098.

Page 187

www.rsisinternational.org

MIC3ST 2025 | International Journal of Research and Innovation in Social Science (IJRISS)

Virtual Conference on Melaka International Social Sciences, Science and Technology 2025

ISSN: 2454-6186 | DOI: 10.47772/IJRISS | Special Issue | Volume IX Issue XXIII October 2025

10. Sulaiman H. and Jamil N. (2014), Information security governance model to enhance zakat information

management in Malaysian zakat institutions, Proceedings of the 6th International Conference on

Information Technology and Multimedia, https://doi.org/10.1109/icimu.2014.7066630.

11. Radzi, N. M., & Kenayathulla, H. B. (2017). Educational Equity for Malaysian Urban Poor Muslim

Students. Malaysian Online Journal of Educational Management, 5(4), 47–62.

https://doi.org/10.22452/mojem.vol5no4.4

12. Saad, A. Y. Q., & Al Foori, A. M. (2020). Zakat and Tax: A Comparative Study in Malaysia.

International Journal of Innovation, Creativity and Change, 10(12), 140–151.

https://www.ijicc.net/images/vol10iss12/101213_Saad_2020_E_R.pdf

13. Zulkifli, M. F., Taha, R., Awang, R. @ N., Mohd Nor, M. N., & Ali, A. (2021). Combating Poverty in

Malaysia: The Role of Zakat. The Journal of Asian Finance, Economics and Business, 8(5), 505–513.

https://doi.org/10.13106/jafeb.2021.vol8.no5.0505

14. Khalil, A. J., Barhoom, A. M., Abu-Nasser, B. S., Musleh, M. M., & Abu-Naser, S. S. (2019). Energy

Efficiency Predicting using Artificial Neural Network. International Journal of Academic Pedagogical

Research (IJAPR), 3(9), 1–8. http://dstore.alazhar.edu.ps/xmlui/handle/123456789/487

15. Al-Mubayyed, O. M., Abu-Nasser, B. S., & Abu-Naser, S. S. (2019). Predicting Overall Car

Performance Using Artificial Neural Network. International Journal of Academic and Applied

Research (IJAAR), 3(1), 1–5. http://dstore.alazhar.edu.ps/xmlui/handle/123456789/127

16. Raptodimos, Y., & Lazakis, I. (2018). Using artificial neural network-self-organising map for data

clustering of marine engine condition monitoring applications. Ships and Offshore Structures, 13(6),

649–656. https://doi.org/10.1080/17445302.2018.1443694