Gerry Firmansyah
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Gerry Firmansyah has a total of 55 co-authors. Here are the top co-authors by number of shared publications:
- Agung Mulyo Widodo
Universitas Esa Unggul
23 shared publications - Budi Tjahjono
Faculty of Computer Science, Universitas Esa Unggul, Indonesia
21 shared publications - Habibullah Akbar
Universitas Esa Unggul
17 shared publications - Nenden Siti Fatonah
Universitas Esa Unggul
8 shared publications - Budi Tjahjono
Teknik Informatika, Fakultas Ilmu Komputer, Universitas Esa Unggul
2 shared publications - Abdullah Qiqi Asmara
Universitas Muhammadiyah Jakarta
1 shared publication
Latest Publications
- Open Access
- 21 December 2025
- Published by State Islamic University of Raden Fatah Palembang in JUSIFO (Jurnal Sistem Informasi)
- Vol. 11 (2) , 133-140
Clustering-Based Identification of Student Support Needs in Higher Education Transition
The transition from secondary to higher education represents a critical phase influenced by both academic readiness and socio-economic conditions. This study proposes a clustering-based approach to identify student support needs during this transition by analyzing multidimensional student profiles. Using secondary data from 1,226 senior high school students, three unsupervised clustering algorithms—K-Means, DBSCAN, and BIRCH—were applied to academic performance and socio-economic variables. Cluster quality was assessed using internal validation metrics, including the Silhouette Score, Davies–Bouldin Index, and Calinski–Harabasz Index. The results indicate that clustering-based methods provide richer insights than traditional rule-based approaches by capturing heterogeneous student profiles and revealing atypical cases. Among the evaluated algorithms, BIRCH demonstrated the most balanced performance in terms of cluster compactness and separation, while K-Means offered stable and interpretable results, and DBSCAN was effective in identifying outliers. Interpreted within the college readiness framework, the identified clusters highlight differentiated student support needs, enabling more targeted and equitable intervention strategies. These findings underscore the potential of educational data mining to support data-driven decision-making in facilitating students’ transition to higher education.
- 21 December 2025
- Published by State Islamic University of Raden Fatah Palembang in JUSIFO (Jurnal Sistem Informasi)
- Vol. 11 (2) , 133-140
Open AccessClustering-Based Identification of Student Support Needs in Higher Education Transition
The transition from secondary to higher education represents a critical phase influenced by both academic readiness and socio-economic conditions. This study proposes a clustering-based approach to identify student support needs during this transition by analyzing multidimensional student profiles. Using secondary data from 1,226 senior high school students, three unsupervised clustering algorithms—K-Means, DBSCAN, and BIRCH—were applied to academic performance and socio-economic variables. Cluster quality was assessed using internal validation metrics, including the Silhouette Score, Davies–Bouldin Index, and Calinski–Harabasz Index. The results indicate that clustering-based methods provide richer insights than traditional rule-based approaches by capturing heterogeneous student profiles and revealing atypical cases. Among the evaluated algorithms, BIRCH demonstrated the most balanced performance in terms of cluster compactness and separation, while K-Means offered stable and interpretable results, and DBSCAN was effective in identifying outliers. Interpreted within the college readiness framework, the identified clusters highlight differentiated student support needs, enabling more targeted and equitable intervention strategies. These findings underscore the potential of educational data mining to support data-driven decision-making in facilitating students’ transition to higher education.
- Open Access
- 5 November 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (11) , 10829-10838
Comprehensive Benchmarking of Message Brokers: Evaluating Performance and Security Metrics for Reliable Messaging Systems
A message broker plays a crucial role in distributed communication systems by ensuring reliable message delivery between services. In practice, choosing a message broker involves not only considering performance aspects such as latency and throughput but also security features like authentication, authorization, and encryption. However, challenges arise when enhancing security layers negatively impacts system performance. This study aims to explore and evaluate four popular message brokers—Apache Kafka, Apache Pulsar, Apache RocketMQ, and RabbitMQ—focusing on performance metrics (latency and throughput) across various security configuration levels (no security, TLS, basic authentication, and advanced authentication). The evaluation is conducted through standardized benchmarking processes to assess the impact of security feature implementation on overall system performance. The results of this research are expected to provide insights into the optimal balance between security and performance that remains acceptable by industry standards. Furthermore, the findings may assist developers and system architects in selecting the most suitable message broker configuration based on scalability requirements and data sensitivity levels.
- 5 November 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (11) , 10829-10838
Open AccessComprehensive Benchmarking of Message Brokers: Evaluating Performance and Security Metrics for Reliable Messaging Systems
A message broker plays a crucial role in distributed communication systems by ensuring reliable message delivery between services. In practice, choosing a message broker involves not only considering performance aspects such as latency and throughput but also security features like authentication, authorization, and encryption. However, challenges arise when enhancing security layers negatively impacts system performance. This study aims to explore and evaluate four popular message brokers—Apache Kafka, Apache Pulsar, Apache RocketMQ, and RabbitMQ—focusing on performance metrics (latency and throughput) across various security configuration levels (no security, TLS, basic authentication, and advanced authentication). The evaluation is conducted through standardized benchmarking processes to assess the impact of security feature implementation on overall system performance. The results of this research are expected to provide insights into the optimal balance between security and performance that remains acceptable by industry standards. Furthermore, the findings may assist developers and system architects in selecting the most suitable message broker configuration based on scalability requirements and data sensitivity levels.
- Open Access
- 5 November 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (11) , 10720-10739
Analisis Kerangka Kerja IT Governance Untuk Memitigasi Risiko Serangan Siber Pada Perusahaan Di Sektor Jasa Keuangan
Cybersecurity has become a critical issue for the financial services sector, which increasingly relies on information technology. Limited resources and inadequate Governance make this sector vulnerable to cyberattacks, which can lead to financial losses, data breaches, and operational disruptions. Therefore, implementing an effective IT Governance Framework is essential to enhance cyber resilience and mitigate attack risks. This study aims to analyze the role of IT Governance in mitigating cyberattack risks for companies in the financial services sector. The research employs a qualitative approach through case studies, interviews, and analysis of relevant documents. The findings indicate that most financial service companies still have low preparedness levels in facing cyber threats. Proper implementation of IT Governance can help improve protection and response to attacks, despite challenges in resources and security awareness.
- 5 November 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (11) , 10720-10739
Open AccessAnalisis Kerangka Kerja IT Governance Untuk Memitigasi Risiko Serangan Siber Pada Perusahaan Di Sektor Jasa Keuangan
Cybersecurity has become a critical issue for the financial services sector, which increasingly relies on information technology. Limited resources and inadequate Governance make this sector vulnerable to cyberattacks, which can lead to financial losses, data breaches, and operational disruptions. Therefore, implementing an effective IT Governance Framework is essential to enhance cyber resilience and mitigate attack risks. This study aims to analyze the role of IT Governance in mitigating cyberattack risks for companies in the financial services sector. The research employs a qualitative approach through case studies, interviews, and analysis of relevant documents. The findings indicate that most financial service companies still have low preparedness levels in facing cyber threats. Proper implementation of IT Governance can help improve protection and response to attacks, despite challenges in resources and security awareness.
- Open Access
- 5 November 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (11) , 10142-10159
Pengembangan Model IT Masterplan untuk Perguruan Tinggi: Studi Kasus Pada Universitas Bakrie dengan Pendekatan TOGAF
Transformasi digital di perguruan tinggi menuntut perencanaan teknologi informasi yang strategis, terstruktur, dan berkelanjutan agar mampu mendukung efisiensi akademik dan administrasi. Penelitian ini bertujuan untuk menyusun IT Masterplan Universitas Bakrie sebagai kerangka strategis pengembangan sistem informasi dan transformasi digital universitas secara holistik. Metode penelitian menggunakan pendekatan kualitatif dengan kombinasi analisis dokumen, wawancara semi-terstruktur, dan survei di lingkungan Universitas Bakrie, termasuk Biro Rektorat, Direktorat TI, dan fakultas-fakultas. Analisis dilakukan dengan mengacu pada kerangka kerja Enterprise Architecture (EA) menggunakan TOGAF Architecture Development Method (ADM) serta analisis kesenjangan (gap analysis) antara kondisi eksisting dan target arsitektur yang diinginkan. Hasil penelitian menunjukkan bahwa kondisi sistem TI Universitas Bakrie masih terfragmentasi, dengan beberapa aplikasi tidak saling terintegrasi dan data yang tersebar. Rekomendasi utama berupa rancangan arsitektur target yang menerapkan model berbasis layanan (service-oriented architecture) dan integrasi melalui middleware untuk mendukung pertukaran data lintas unit secara real-time. Selain itu, disusun pula desain data warehouse terpusat serta mekanisme tata kelola data (data governance) untuk meningkatkan akurasi dan analitik lintas fungsi. Implementasi IT Masterplan ini diharapkan mampu meningkatkan efisiensi pengelolaan sumber daya TI, mengurangi redundansi sistem, memperkuat interoperabilitas aplikasi, serta mendorong kelincahan digital universitas dalam menghadapi era transformasi pendidikan tinggi.
- 5 November 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (11) , 10142-10159
Open AccessPengembangan Model IT Masterplan untuk Perguruan Tinggi: Studi Kasus Pada Universitas Bakrie dengan Pendekatan TOGAF
Transformasi digital di perguruan tinggi menuntut perencanaan teknologi informasi yang strategis, terstruktur, dan berkelanjutan agar mampu mendukung efisiensi akademik dan administrasi. Penelitian ini bertujuan untuk menyusun IT Masterplan Universitas Bakrie sebagai kerangka strategis pengembangan sistem informasi dan transformasi digital universitas secara holistik. Metode penelitian menggunakan pendekatan kualitatif dengan kombinasi analisis dokumen, wawancara semi-terstruktur, dan survei di lingkungan Universitas Bakrie, termasuk Biro Rektorat, Direktorat TI, dan fakultas-fakultas. Analisis dilakukan dengan mengacu pada kerangka kerja Enterprise Architecture (EA) menggunakan TOGAF Architecture Development Method (ADM) serta analisis kesenjangan (gap analysis) antara kondisi eksisting dan target arsitektur yang diinginkan. Hasil penelitian menunjukkan bahwa kondisi sistem TI Universitas Bakrie masih terfragmentasi, dengan beberapa aplikasi tidak saling terintegrasi dan data yang tersebar. Rekomendasi utama berupa rancangan arsitektur target yang menerapkan model berbasis layanan (service-oriented architecture) dan integrasi melalui middleware untuk mendukung pertukaran data lintas unit secara real-time. Selain itu, disusun pula desain data warehouse terpusat serta mekanisme tata kelola data (data governance) untuk meningkatkan akurasi dan analitik lintas fungsi. Implementasi IT Masterplan ini diharapkan mampu meningkatkan efisiensi pengelolaan sumber daya TI, mengurangi redundansi sistem, memperkuat interoperabilitas aplikasi, serta mendorong kelincahan digital universitas dalam menghadapi era transformasi pendidikan tinggi.
- Open Access
- 3 September 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (9) , 8896-8905.
Evaluasi dan Optimasi Kinerja MySQL Master-Slave dengan Metode Kuantitatif pada Database Pemohon Tes Psikologi SIM PT XYZ pada POLDA METRO JAYA
The development of information technology and cloud computing has enabled organizations to manage large-scale data efficiently. PT XYZ, which is engaged in psychological testing for Driving Licenses (SIM), uses a web-based system with a MySQL database that has implemented master-slave replication. However, as the data volume increases to 4,000-5,000 entries per day, the system experiences performance constraints, especially in the speed of read and write queries. This study aims to optimize the performance of the MySQL database by adjusting the server configuration and specifications to improve system efficiency. The test results show that server specification settings, including processor speed, memory size, and replication configuration, play an important role in improving system performance. By adjusting the master and slave server configurations, the system shows a significant increase in database response time and operational efficiency. This optimization is expected to be a reference in the implementation and management of large-scale databases using MySQL replication.
- 3 September 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (9) , 8896-8905.
Open AccessEvaluasi dan Optimasi Kinerja MySQL Master-Slave dengan Metode Kuantitatif pada Database Pemohon Tes Psikologi SIM PT XYZ pada POLDA METRO JAYA
The development of information technology and cloud computing has enabled organizations to manage large-scale data efficiently. PT XYZ, which is engaged in psychological testing for Driving Licenses (SIM), uses a web-based system with a MySQL database that has implemented master-slave replication. However, as the data volume increases to 4,000-5,000 entries per day, the system experiences performance constraints, especially in the speed of read and write queries. This study aims to optimize the performance of the MySQL database by adjusting the server configuration and specifications to improve system efficiency. The test results show that server specification settings, including processor speed, memory size, and replication configuration, play an important role in improving system performance. By adjusting the master and slave server configurations, the system shows a significant increase in database response time and operational efficiency. This optimization is expected to be a reference in the implementation and management of large-scale databases using MySQL replication.
- Open Access
- 1 August 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (8) , 7958-7968.
Analisis dan Design Knowledge Management System pada PT XYZ dengan Menggunakan Metode Tiwana
Dalam dunia bisnis, keberlangsungan dan pertumbuhan suatu perusahaan sangat dipengaruhi oleh kualitas sumber daya manusia (SDM) atau karyawan yang dimiliki, karyawan yang kompeten dan memiliki wawasan pengetahuan yang luas dianggap sebagai aset penting yang dapat menjadi pembeda utama antara perusahaan dengan para pesaingnya. Maka dari itu, perusahaan biasanya berinisiatif untuk melakukan berbagai macam program pelatihan dan pengembangan yang bertujuan untuk meningkatkan kemampuan dan pengetahuan karyawannya secara berkelanjutan, hal lain yang dapat dilakukan oleh perusahaan adalah dengan mengelola pengetahuan dari karyawannya melalui sebuah tools yaitu Knowledge Management System (KMS). Sebagai perusahaan konsultan properti terkemuka dengan jaringan global yang luas, PT XYZ memiliki komitmen untuk dapat mengembangkan pengetahuan karyawannya, PT XYZ selalu berupaya untuk menciptakan lingkungan kerja yang mendukung pertumbuhan pengetahuan, terutama bagi karyawan di bidang Information and Technology (IT). Penelitian ini menggunakan metode Tiwana yang dilakukan hingga tahapan ke-6 dari 10 tahapan yang ada, dengan tujuan penelitian untuk menghasilkan blueprint sebagai landasan untuk merancang KMS yang sesuai dengan kebutuhan perusahaan. Adapun hasil dari penelitian ini adalah sebuah blueprint rancangan KMS.
- 1 August 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (8) , 7958-7968.
Open AccessAnalisis dan Design Knowledge Management System pada PT XYZ dengan Menggunakan Metode Tiwana
Dalam dunia bisnis, keberlangsungan dan pertumbuhan suatu perusahaan sangat dipengaruhi oleh kualitas sumber daya manusia (SDM) atau karyawan yang dimiliki, karyawan yang kompeten dan memiliki wawasan pengetahuan yang luas dianggap sebagai aset penting yang dapat menjadi pembeda utama antara perusahaan dengan para pesaingnya. Maka dari itu, perusahaan biasanya berinisiatif untuk melakukan berbagai macam program pelatihan dan pengembangan yang bertujuan untuk meningkatkan kemampuan dan pengetahuan karyawannya secara berkelanjutan, hal lain yang dapat dilakukan oleh perusahaan adalah dengan mengelola pengetahuan dari karyawannya melalui sebuah tools yaitu Knowledge Management System (KMS). Sebagai perusahaan konsultan properti terkemuka dengan jaringan global yang luas, PT XYZ memiliki komitmen untuk dapat mengembangkan pengetahuan karyawannya, PT XYZ selalu berupaya untuk menciptakan lingkungan kerja yang mendukung pertumbuhan pengetahuan, terutama bagi karyawan di bidang Information and Technology (IT). Penelitian ini menggunakan metode Tiwana yang dilakukan hingga tahapan ke-6 dari 10 tahapan yang ada, dengan tujuan penelitian untuk menghasilkan blueprint sebagai landasan untuk merancang KMS yang sesuai dengan kebutuhan perusahaan. Adapun hasil dari penelitian ini adalah sebuah blueprint rancangan KMS.
- Open Access
- 1 August 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (8) , 8303-8318
Audit Tata Kelola Teknologi Informasi Menggunakan Framework COBIT 2019 Pada Rumah Sakit Medika Dramaga
Penelitian ini bertujuan untuk mengevaluasi tata kelola teknologi informasi (TI) di Rumah Sakit Medika Dramaga menggunakan kerangka kerja COBIT 2019. Hasil audit menunjukkan bahwa meskipun beberapa aspek tata kelola TI telah mencapai tingkat tertentu, masih terdapat kesenjangan antara kondisi saat ini dan harapan yang diinginkan. Pada domain APO12 (Pengelolaan Risiko), tingkat kemampuan mencapai 87% pada level 2 dan 73% pada level 3, namun tidak ada pencapaian pada level 4 dan 5. Sementara itu, pada domain APO13 (Pengelolaan Keamanan Informasi), tingkat kemampuan hanya mencapai 82% pada level 2 tanpa pencapaian pada level yang lebih tinggi. Kesenjangan ini menunjukkan bahwa pengelolaan risiko dan keamanan informasi masih memerlukan peningkatan signifikan untuk mencapai standar yang diharapkan. Berdasarkan temuan tersebut, penelitian ini memberikan beberapa rekomendasi, termasuk evaluasi kebijakan manajemen risiko, implementasi teknologi pendukung, pelatihan SDM, dan pengembangan strategi jangka panjang untuk meningkatkan tata kelola TI. Dengan menerapkan rekomendasi ini, diharapkan Rumah Sakit Medika Dramaga dapat meningkatkan keamanan dan keandalan sistem informasi serta meminimalkan risiko kebocoran data.
- 1 August 2025
- Published by Riviera Publishing in Jurnal Locus Penelitian dan Pengabdian
- Vol. 4 (8) , 8303-8318
Open AccessAudit Tata Kelola Teknologi Informasi Menggunakan Framework COBIT 2019 Pada Rumah Sakit Medika Dramaga
Penelitian ini bertujuan untuk mengevaluasi tata kelola teknologi informasi (TI) di Rumah Sakit Medika Dramaga menggunakan kerangka kerja COBIT 2019. Hasil audit menunjukkan bahwa meskipun beberapa aspek tata kelola TI telah mencapai tingkat tertentu, masih terdapat kesenjangan antara kondisi saat ini dan harapan yang diinginkan. Pada domain APO12 (Pengelolaan Risiko), tingkat kemampuan mencapai 87% pada level 2 dan 73% pada level 3, namun tidak ada pencapaian pada level 4 dan 5. Sementara itu, pada domain APO13 (Pengelolaan Keamanan Informasi), tingkat kemampuan hanya mencapai 82% pada level 2 tanpa pencapaian pada level yang lebih tinggi. Kesenjangan ini menunjukkan bahwa pengelolaan risiko dan keamanan informasi masih memerlukan peningkatan signifikan untuk mencapai standar yang diharapkan. Berdasarkan temuan tersebut, penelitian ini memberikan beberapa rekomendasi, termasuk evaluasi kebijakan manajemen risiko, implementasi teknologi pendukung, pelatihan SDM, dan pengembangan strategi jangka panjang untuk meningkatkan tata kelola TI. Dengan menerapkan rekomendasi ini, diharapkan Rumah Sakit Medika Dramaga dapat meningkatkan keamanan dan keandalan sistem informasi serta meminimalkan risiko kebocoran data.
- Open Access
- 11 April 2025
- Published by Politeknik Ganesha in sinkron
- Vol. 9 (2) , 841-848
Metaheuristic-Optimized SVM for Stunting Risk Detection in Pregnancy
Stunting is a chronic growth disorder that originates during pregnancy, making early risk detection crucial for effective prevention and long-term child development. This study introduces a stunting risk prediction model based on urine testing, employing a Support Vector Machine (SVM) algorithm enhanced through metaheuristic optimization. Three metaheuristic algorithms—Grey Wolf Optimizer (GWO), Simulated Annealing (SA), and Firefly Algorithm (FA)—were utilized to fine-tune the SVM hyperparameters (C and gamma). Clinical urine samples collected from pregnant women served as the dataset for model training and validation. The results indicate that the SVM model optimized using GWO achieved the highest prediction accuracy at 94.15%, outperforming both the default SVM (88.46%) and the models optimized using SA (94.12%) and FA (85.71%). Additionally, significant improvements were observed in precision, recall, and F1-score metrics, affirming the effectiveness of metaheuristic tuning in enhancing classification performance. These findings highlight the potential of integrating metaheuristic algorithms with SVM for robust medical prediction tasks, especially in the early detection of stunting risks. The proposed model offers a promising and non-invasive diagnostic approach that can be implemented in prenatal care settings, enabling timely interventions to mitigate stunting and improve maternal and child health outcomes.
- 11 April 2025
- Published by Politeknik Ganesha in sinkron
- Vol. 9 (2) , 841-848
Open AccessMetaheuristic-Optimized SVM for Stunting Risk Detection in Pregnancy
Stunting is a chronic growth disorder that originates during pregnancy, making early risk detection crucial for effective prevention and long-term child development. This study introduces a stunting risk prediction model based on urine testing, employing a Support Vector Machine (SVM) algorithm enhanced through metaheuristic optimization. Three metaheuristic algorithms—Grey Wolf Optimizer (GWO), Simulated Annealing (SA), and Firefly Algorithm (FA)—were utilized to fine-tune the SVM hyperparameters (C and gamma). Clinical urine samples collected from pregnant women served as the dataset for model training and validation. The results indicate that the SVM model optimized using GWO achieved the highest prediction accuracy at 94.15%, outperforming both the default SVM (88.46%) and the models optimized using SA (94.12%) and FA (85.71%). Additionally, significant improvements were observed in precision, recall, and F1-score metrics, affirming the effectiveness of metaheuristic tuning in enhancing classification performance. These findings highlight the potential of integrating metaheuristic algorithms with SVM for robust medical prediction tasks, especially in the early detection of stunting risks. The proposed model offers a promising and non-invasive diagnostic approach that can be implemented in prenatal care settings, enabling timely interventions to mitigate stunting and improve maternal and child health outcomes.
- Open Access
- 11 April 2025
- Published by Politeknik Ganesha in sinkron
- Vol. 9 (2) , 822-831
Loan Repayment Prediction Using XGBoost and Neural Network in Japan's Technical Internship Training
Delayed repayment of financial aid among participants in Japan’s Technical Internship Training Program presents challenges for training institutions in managing funds efficiently. To address this issue, this study aims to compare the performance of two machine learning models: Extreme Gradient Boosting (XGBoost) and Multi-Layer Perceptron (MLP) in predicting the likelihood of delayed loan repayments. The research begins with data preprocessing, including handling missing values, normalization, and feature selection based on a correlation threshold of 0.06, where features with absolute correlation values below this threshold are excluded. Three models are tested: XGBoost Default, XGBoost optimized using GridSearchCV, and MLP. These models are evaluated using performance metrics such as accuracy, precision, recall, F1-score, and ROC-AUC. The XGBoost Default model achieves the highest accuracy at 95% and precision of 95%, although its recall is slightly lower at 83%. Tuning XGBoost improves recall to 84%, albeit with a marginal reduction in accuracy to 94%. In contrast, the MLP model demonstrates the lowest performance, with an accuracy of 92% and recall of 74%, indicating limitations in identifying delayed repayments. XGBoost also outperforms MLP in terms of ROC-AUC, scoring 91% compared to MLP’s 86%. These findings suggest that XGBoost is the more effective model for this predictive task. The results have practical implications for training institutions, enabling better participant selection, reducing repayment delays, and supporting more effective financial aid management.
- 11 April 2025
- Published by Politeknik Ganesha in sinkron
- Vol. 9 (2) , 822-831
Open AccessLoan Repayment Prediction Using XGBoost and Neural Network in Japan's Technical Internship Training
Delayed repayment of financial aid among participants in Japan’s Technical Internship Training Program presents challenges for training institutions in managing funds efficiently. To address this issue, this study aims to compare the performance of two machine learning models: Extreme Gradient Boosting (XGBoost) and Multi-Layer Perceptron (MLP) in predicting the likelihood of delayed loan repayments. The research begins with data preprocessing, including handling missing values, normalization, and feature selection based on a correlation threshold of 0.06, where features with absolute correlation values below this threshold are excluded. Three models are tested: XGBoost Default, XGBoost optimized using GridSearchCV, and MLP. These models are evaluated using performance metrics such as accuracy, precision, recall, F1-score, and ROC-AUC. The XGBoost Default model achieves the highest accuracy at 95% and precision of 95%, although its recall is slightly lower at 83%. Tuning XGBoost improves recall to 84%, albeit with a marginal reduction in accuracy to 94%. In contrast, the MLP model demonstrates the lowest performance, with an accuracy of 92% and recall of 74%, indicating limitations in identifying delayed repayments. XGBoost also outperforms MLP in terms of ROC-AUC, scoring 91% compared to MLP’s 86%. These findings suggest that XGBoost is the more effective model for this predictive task. The results have practical implications for training institutions, enabling better participant selection, reducing repayment delays, and supporting more effective financial aid management.
- 17 March 2025
- Published by Springer Nature
- p. 168-180
Security Benefits in Small Medium Business Enterprise Through Cloud Adoption
- 17 March 2025
- Published by Springer Nature
- p. 168-180
Security Benefits in Small Medium Business Enterprise Through Cloud Adoption