Claim Missing Document
Check
Articles

Decision-making model for cadet selection using the AHP TOPSIS method Tazky Tsany; Jonson Manurung; M. Azhar Prabukusumo
Jurnal Mandiri IT Vol. 14 No. 3 (2026): Jan: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v14i3.486

Abstract

Cadet selection in defense institutions requires a comprehensive assessment process because it must cover the academic, psychological, health, physical, and ideological integrity aspects of prospective participants. This multidimensional complexity poses challenges in producing decisions that are objective, consistent, and free from assessor bias. Therefore, a quantitative approach-based evaluation model is needed that can integrate all assessment components in a measurable manner. This study developed a cadet selection decision-making model using a combination of the Analytical Hierarchy Process (AHP) and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) methods. AHP is used to determine the weight of importance of the seven main criteria: Academic Potential Test, Academic Interview, Psychological Test, Ideological Mental Test, Ideological Mental Interview, Health Test, and Physical Test, while TOPSIS is used to determine the ranking of candidates based on their proximity to the ideal profile of a cadet. The results of the study show that the integration of AHP–TOPSIS is able to provide evaluation results that are more objective, transparent, and accountable than conventional assessments. In addition to formulating a selection model, this study also discusses alternative methods in multi-criteria decision making as material for developing a selection system in the future. Overall, this model is expected to become a scientific basis for defense institutions in improving the quality and accuracy of the cadet selection process.
Mixed integer linear programming for cadet dormitory placement at Indonesia Defense University I Made Aditya Pradhana Putra; Jonson Manurung; Hondor Saragih
Jurnal Mandiri IT Vol. 14 No. 3 (2026): Jan: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v14i3.487

Abstract

Cadet dormitory placement at Indonesian Defense University was currently performed manually by administrative staff, resulting in potential inefficiencies in room assignments regarding walking distance, study program cohesion, and cadet preferences. This research developed a Mixed Integer Linear Programming (MILP) optimization model to automate and improve the dormitory assignment process for military education institutions. The general framework addresses 1,550 cadets distributed across four cohorts and 13 study programs in   dormitory buildings with standardized configurations (3 floors, 25 rooms per floor, 2 cadets per room). The MILP model incorporated three objectives: minimizing total walking distance to academic facilities, maximizing study program cohesion by concentrating programs within specific floors, and maximizing cadet floor preference satisfaction. The model was formulated with configurable weight parameters (w₁, w₂, w₃) enabling administrators to balance competing objectives according to institutional priorities. A validation case study with 38 male cadets from two study programs demonstrated computational feasibility, with the CBC solver achieving optimal solutions in 0.34 seconds (strict constraint approach) and 0.11 seconds (maximum occupancy approach) on standard desktop hardware, both with 0.00% MIP gap confirming proven optimality. The validation study compared two policy approaches: strict constraint enforcement achieving 95% room occupancy with 20 rooms, and maximum space utilization achieving 100% occupancy with 19 rooms. This research contributed the first application of MILP optimization to military education dormitory management in Indonesia, providing a scalable framework with empirical validation for computational tractability and a replicable methodology for resource allocation optimization in defense institutions.
Mapping monthly consumer purchasing patterns at the UNHAN RI Cooperative using time series analysis and LSTM Miranda Bintang Maharani Sigalingging; M. Azhar Prabukusumo Prabukusumo; Jonson Manurung
Jurnal Mandiri IT Vol. 14 No. 3 (2026): Jan: Computer Science and Field
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mandiri.v14i3.488

Abstract

This study investigated the monthly purchasing patterns of consumers at Koperasi Unhan RI and developed forecasting models to support data-driven inventory and procurement planning. Historical cooperative sales data from 2020–2024 were analyzed using time series decomposition, autocorrelation analysis, ARIMA modeling, and a Long Short-Term Memory (LSTM) neural network. The analysis revealed a clear upward trend and strong annual seasonality, with consistent demand peaks occurring in December. The ARIMA model achieved significantly lower prediction errors than the LSTM model and successfully captured both trend and seasonal components. A 12-month forecast for 2025 was then generated to support operational decision-making. The forecasting results provide practical managerial insights for cooperative management, particularly in optimizing inventory levels, scheduling procurement, and anticipating seasonal demand fluctuations. The novelty of this study lies in the comparative application of classical time-series and deep learning approaches within a cooperative context using limited historical data, demonstrating that ARIMA remains a robust and interpretable solution for small to medium-sized cooperative environments. This research concludes that time series analysis combined with ARIMA forecasting effectively mapped consumer purchasing patterns and produced actionable demand predictions for the subsequent year.
Cyber threat detection on social media using indoBERT and sentiment analysis Bagus Hendra Saputra; Jonson Manurung; Baringin Sianipar; R. Fanry Siahaan
Journal of Defense Technology and Engineering Vol. 2 No. 1 (2026): July, Journal of Defense Technology and Engineering
Publisher : Fakultas Teknik dan Teknologi Pertahanan, Universitas Pertahanan Republik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The rapid growth of Indonesian social media has increased the spread of cyber threat-related content, creating significant challenges for digital security monitoring due to the informal language, code-switching, and sentiment-rich expressions commonly used in online communication. Existing detection approaches, particularly those based on multilingual or English-centric language models, often fail to capture the linguistic characteristics of Indonesian text effectively. This study aims to develop an accurate cyber threat detection model by fine-tuning IndoBERT, a transformer-based language model pretrained on a large-scale Indonesian corpus, for binary Threat and Non-Threat classification. The model was trained and evaluated using the Tweet ID Sentiment Dataset containing 10,800 annotated tweets, which were partitioned into training, validation, and test sets, and its performance was compared with four baseline methods: SVM with TF-IDF features, CNN with FastText embeddings, BiLSTM with Word2Vec representations, and multilingual BERT. Experimental results demonstrate that the proposed IndoBERT model achieved the best performance, obtaining an accuracy of 0.9389, a macro-F1 score of 0.9292, and a Threat-class recall of 0.9486, consistently outperforming all baseline models. The novelty of this study lies in demonstrating the effectiveness of a monolingual Indonesian pretrained transformer for cyber threat detection, highlighting the importance of language-specific contextual representations in improving classification performance. These findings indicate that the proposed approach provides a robust and practical solution for automated cyber threat detection, supporting early warning systems and digital security monitoring in Indonesian social media environments. Future work will investigate multiclass cyber threat categorization and cross-platform evaluation to improve model generalizability in real-world applications.
An explainable deep learning framework for autonomous drone-based military object detection using vision transformers Jontinus Manullang; R. Fanry Siahaan; Sutrisno Situmorang; Jonson Manurung
Journal of Defense Technology and Engineering Vol. 2 No. 1 (2026): July, Journal of Defense Technology and Engineering
Publisher : Fakultas Teknik dan Teknologi Pertahanan, Universitas Pertahanan Republik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The rapid expansion of overhead imagery and unmanned aerial sensing has intensified the need for accurate, interpretable, and computationally adaptable object detection models in security-sensitive remote-sensing environments. The main research problem addressed in this study is the limited transparency of deep object detection models when identifying small, dense, and visually ambiguous objects in aerial or satellite imagery. This study aims to design an explainable deep learning framework that combines Vision Transformer representation learning, transfer learning, and Grad-CAM-based visual explanation for autonomous drone-based military object detection using the xView dataset. The proposed research process includes dataset preparation, image tiling, bounding-box conversion, normalization, transfer learning from an ImageNet-pretrained ViT backbone, transformer-based feature extraction, Grad-CAM visualization, and evaluation using accuracy, precision, recall, F1-score, IoU, AP, and mAP. The dataset used in this manuscript is the xView dataset available through a Kaggle mirror and originally introduced as a large-scale overhead imagery benchmark containing more than one million labeled object instances across 60 categories. Experimental results indicate that the proposed ViT + Grad-CAM framework achieves 0.823 accuracy, 0.814 macro precision, 0.806 macro recall, 0.810 macro F1-score, and 0.802 mAP@0.5 on a five-class defense-relevant subset, outperforming SSD, Faster R-CNN, and YOLOv8 baselines in the reported comparison. The study concludes that combining transformer-based global context modeling with explainability can improve both detection performance and model auditability in overhead imagery, although real deployment requires full-dataset validation, edge-device optimization, robustness testing, and strict human oversight for ethical use.
Co-Authors Adam Mardamsyah Adha, Rochedi Idul Agus Firmansyah Agustina Simangunsong Ajeng Hidayati Al Hashim, Safa Ayoub Andri Budiman, Mohammad anindito anindito Arnof, Muhammad Rafi Bagus Hendra Saputra Bagus Hendra Saputra Baringin Sianipar Barus, Nadela Bosker Sinaga Bosker Sinaga Bosker Sinaga, Bosker Sinaga Br Sitepu, Siska Feronika Br Tarigan, Nera Mayana Cloudio Egi Primsa Ginting Dhaifullah, Rendi Hanif Erika Novianti Eryan Ahmad Firdaus Eryan Ahmad Firdaus Eryan Ahmad Firdaus Febrian Wahyu Christanto Ferdinand Tharorogo Wau Firdaus Laia Firdaus Situmorang Frainskoy Rio Naibaho Hanan, Rohman Ali Hardy Priyatno Ambarita Hasugian , Paska Marto Hengki Tamando Sihotang Hidayati, Ajeng Hoga Saragih Hondor Saragih Hondor Saragih I Made Aditya Pradhana Putra Jeremia Paska Sinaga Jeremia Paskah Sinaga Jeremia Sinaga Johanes Perdamenta Sembiring Jontinus Manullang Kadin Darlianto Tinambunan Kanur L. P. Situmorang Kartika, Amelia Khaerul Imam Phatoni Logaraj Logaraj Logaraj, Logaraj Lumban Tobing, Ruben M Azhar Prabukusumo M Azhar Prabukusumo M. Azhar Prabukusumo M. Azhar Prabukusumo Prabukusumo Maria Siahaan Maya Theresia Br. Barus Maya Theresia Br. Barus Merlin Helentina Napitupulu Mina Kumari Miranda Bintang Maharani Sigalingging Muhammad Azhar Prabukusumo Muhammad Sulthan Nasyira Muthmainnah, Ihmatull Nick Holson M. Silalahi Nisrina Labiba Sarwoko Nuriansyah, Agam Pandiangan, Boyner Poltak Sihombing Prabukusumo, Muhammad Azhar Putra, Muhammad Ridho Alghifari R. Fanry Siahaan Ram, Uzitha Ramen, Sethu Rinaldy Chaniago Rizqullah Aryaputra Piliang Rochedi Idul Adha Sawaluddin Sawaluddin, Sawaluddin Sethu Ramen Sethu Ramen, Sethu Ramen Sibarani, Harpingka Sidiq, Maulana Sihombing, Agus Putra Emas Sihotang, Amran Silalahi, Monalisa Hotmauli Simangunsong, Humala Sinaga, Ryan Fahlepy Sri Kumala Sari Sutrisno Situmorang Tambunan, Vinaldo Tazky Tsany Vernando, Deden