cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Naety
Contact Email
jurnalmedicom@iocscience.org
Phone
+6281381251442
Journal Mail Official
jurnalmedicom@iocscience.org
Editorial Address
Perumahan Romeby Lestari Blok C, No C14 Deliserdang, Sumatera Utara, Indonesia
Location
Unknown,
Unknown
INDONESIA
Jurnal Teknik Informatika C.I.T. Medicom
Published by Institute of Computer Science
ISSN : 23378646     EISSN : 2721561X     DOI : -
Core Subject : Science,
Computer Science & IT
The Jurnal Teknik Informatika C.I.T a scientific journal of Decision support sistem , expert system and artificial inteligens which includes scholarly writings on pure research and applied research in the field of information systems and information technology as well as a review-general review of the development of the theory, methods, and related applied sciences.
Arjuna Subject : Ilmu Komputer (semua kategori) - Kecerdasan Buatan
Articles 130 Documents
 9   10   11   12   13 
Real Time Pill Counting on Low Power Device: A YOLOv5 Pipeline with Confidence Thresholding and NMS A, Galih Prakoso Rizky; Widyastuti, Rifka
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 5 (2025): November : Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/cit.Vol17.2025.1286.pp225-241

Abstract

Manual pill counting is still commonly performed in healthcare facilities and pharmacies, but this method is vulnerable to human error and requires significant processing time. This study develops an automatic pill counting pipeline using the YOLOv5 deep learning model, optimized for low-power devices such as Raspberry Pi, Orange Pi, and Jetson Nano. Unlike earlier techniques that depend on conventional retrieval or machine-learning approaches, this pipeline integrates real-time object detection with customized confidence thresholding and Non-Maximum Suppression (NMS), enabling high accuracy and fast performance on edge hardware with limited resources. The development process includes collecting and annotating a dataset of pill images with variations in shape, color, and orientation, followed by training YOLOv5 using optimized parameters. A simple webcam is used as the input device, and system performance is evaluated under different lighting and background conditions. Experimental results show that the model achieves 98% precision, 88% recall, 95% mAP@0.5, and 67% mAP@0.5:0.95, with an average inference speed of around 15 milliseconds per image. Tests on ten pill-counting scenarios under optimal lighting demonstrate strong performance, with only minor discrepancies in dense cases involving 50 and 127 pills, producing accuracies of 98% and 99.21%. These results indicate that the optimized YOLOv5 pipeline provides fast and accurate real-time pill counting on low-power devices. Future work will enhance robustness to lighting variations, validate using external datasets, and incorporate color and shape feature analysis to improve performance in challenging scenarios.
A Comprehensive Review of Modern Machine Learning Architectures: From Statistical Models to Adaptive Intelligent Computing Systems Prakoso Rizky A, Galih; Situmorang, Rohani
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 1 (2025): March: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

  This research presents a comprehensive review of the evolution, performance, and limitations of modern machine learning architectures, spanning from classical statistical models to advanced adaptive intelligent computing systems. By systematically comparing diverse architectural families including linear models, tree-based learners, convolutional neural networks (CNNs), recurrent neural networks (RNNs), Transformers, and emerging adaptive systems the study evaluates their computational complexity, training efficiency, scalability, data requirements, interpretability, robustness, and adaptability. The findings reveal that while traditional models remain valuable for their simplicity and transparency, deep learning and Transformer-based architectures significantly outperform earlier methods in handling large-scale, high-dimensional, and unstructured data. However, these performance gains come with notable challenges, including high computational and energy costs, adversarial vulnerability, data bias, lack of explainability, and difficulties in deployment on resource-limited devices. The study also compares current results with key findings from the past decade, highlighting both continuities and major advancements in model capabilities, scalability, and reliability. Overall, the research contributes an integrated framework that synthesizes technical, ethical, and practical considerations, offering deeper insights into the strengths, limitations, and future directions of modern machine learning architectures. The study underscores the need for more interpretable, energy-efficient, and ethically aligned AI systems to support responsible and sustainable technological development.
A Probabilistic Decision Model for AI-Driven Optimization in Highly Complex Systems Riandari, Fristi; Panjaitan, Firta Sari
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 1 (2025): March: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

Highly complex systems such as smart grids, autonomous transportation networks, and large-scale supply chains present significant challenges for optimization due to high dimensionality, nonlinear interactions, and pervasive uncertainty. Traditional deterministic models often fail under dynamic conditions, while many AI-based approaches lack robustness and stability when confronted with noisy or incomplete data. Addressing these issues, this study proposes a probabilistic decision model designed to enhance AI-driven optimization in uncertain and rapidly changing environments. The model integrates probabilistic graphical structures, Bayesian inference, and AI-based optimization techniques to quantify uncertainty and support adaptive decision-making. Experimental evaluations were conducted using a combination of synthetic datasets, simulation environments, and benchmark scenarios representative of real-world complex systems. Results show that the proposed model achieves significantly higher decision accuracy, improved stability under noisy conditions, and more efficient performance in high-dimensional settings compared with classical optimization, reinforcement learning, and standard probabilistic approaches. The model consistently reduces uncertainty and delivers robust, reliable solutions across a wide range of test conditions.The study presents a scalable, interpretable, and highly effective framework for uncertainty-aware optimization. Its strong performance and generalizability highlight its potential for deployment in critical real-world applications where reliability, safety, and adaptability are essential.
A Probabilistic Decision Model for AI-Driven Optimization in Highly Complex Stochastic Mixed-Integer Nonlinear Programming (MINLP) Systems Sihotang, Hengki Tamando; Simbolon, Roma Sinta
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 1 (2025): March: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

Highly complex systems present significant challenges for optimization, particularly when operating under uncertainty, high dimensionality, and dynamic environmental conditions. This study proposes a probabilistic decision model designed to enhance AI-driven optimization by integrating uncertainty quantification, adaptive decision mechanisms, and robust probabilistic reasoning. The methodology combines probabilistic modeling with machine learning techniques and is evaluated through a series of controlled experimental scenarios that simulate real-world complexity and noise. The results indicate substantial improvements in decision accuracy, solution stability, and robustness compared to traditional deterministic and heuristic-based optimization methods. The model consistently maintains high performance despite uncertain inputs and fluctuating system parameters, demonstrating its reliability in environments where conventional approaches tend to degrade. Theoretical analysis further validates the model’s feasibility and guarantees performance consistency under uncertainty. Overall, this research contributes a scalable and resilient decision-making framework capable of addressing the limitations of existing optimization models, offering significant potential for broad application in AI-driven complex systems.
A Unified Hybrid AHP, Utility, TOPSIS Decision Model for Enhancing Ranking Reliability in Complex Multi-Criteria Problems Sihotang, Jonhariono; Batubara, Juliana
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 1 (2025): March: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

This study proposes a unified mathematical framework that integrates the Analytic Hierarchy Process (AHP), the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), and Utility Theory to enhance multi-criteria decision-making (MCDM) in complex environments. While AHP provides a structured mechanism for deriving criterion weights, TOPSIS offers an effective geometric ranking approach, and Utility Theory captures nonlinear preferences and risk attitudes. However, these methods often operate independently, resulting in inconsistent rankings and incomplete representation of decision-maker behavior. The proposed framework bridges these gaps by combining AHP-derived weights, utility-transformed criterion values, and TOPSIS proximity measures into an integrated decision function. A numerical case study illustrates the full application of the model, including weight calculation, utility transformation, ideal-solution analysis, and composite scoring. Results show that the unified model produces more stable and discriminative rankings than pure AHP, pure TOPSIS, or pure Utility Theory. Sensitivity and robustness analyses further demonstrate that the integrated approach maintains ranking consistency under variations in weights, normalization methods, and utility parameters. Comparative validation using Spearman correlation confirms strong agreement with established methods while improving resilience to uncertainty. Overall, this research contributes a comprehensive and theoretically grounded MCDM framework that better reflects human judgment, strengthens ranking reliability, and is adaptable to diverse decision contexts. The unified model offers a powerful tool for practitioners and researchers seeking more accurate and robust decision support in multi-criteria environments.
A Federated Multimodal Learning Framework for Privacy-Preserving Intelligent Computing in Large-Scale IoT Ecosystems Veskardin, Lianora; Threyn, Cassandra R.
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 1 (2025): March: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

The rapid expansion of large-scale Internet of Things (IoT) ecosystems has generated massive volumes of heterogeneous multimodal data, creating new challenges related to scalability, data integration, privacy protection, and real-time intelligence. Traditional centralized learning architectures struggle with communication bottlenecks, privacy regulations, and the complexity of processing diverse data modalities such as sensor signals, audio, video, text, and location streams. Although federated learning (FL) provides a decentralized alternative, existing FL models remain limited in handling multimodal inputs, managing non-IID data distributions, and ensuring strong resilience to adversarial threats. This study proposes a Federated Multimodal Learning Framework that combines probabilistic representation encoding, hierarchical mixture-of-experts fusion, cross-modal consistency regularization, and communication-efficient update scheduling. The framework enables distributed IoT devices to collaboratively learn multimodal representations without sharing raw data, thereby maintaining compliance with GDPR, HIPAA, and other privacy legislation. A probabilistic multimodal embedding mechanism reduces information leakage while supporting dynamic and reliable cross-modal interactions, even under missing or imbalanced modality conditions. Experimental results show that the proposed framework significantly outperforms existing multimodal FL approaches. It achieves higher model accuracy, reduces communication costs by 40-70%, maintains strong privacy protection with minimal performance degradation, and demonstrates enhanced robustness against adversarial attacks. Furthermore, the model provides superior multimodal fusion quality, effectively aligning heterogeneous data streams within federated constraints. Overall, this research delivers a scalable, privacy-preserving, and highly adaptive solution for intelligent computing in modern IoT environments, offering a stronger foundation for real-world applications in smart cities, industrial automation, healthcare monitoring, and next-generation distributed AI systems.
Analyzing the Limitations of Conventional Machine Learning Models in Handling Large-Scale and Heterogeneous Data Prakoso Rizky A, Galih; Situmorang, Rohani
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 2 (2025): May: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

The rapid growth of data volume, dimensionality, and heterogeneity has challenged the effectiveness of conventional machine learning models, which were originally designed for smaller and more homogeneous datasets. This study analyzes the structural and computational limitations of traditional models such as Logistic Regression, Naïve Bayes, Decision Trees, and Support Vector Machines in handling large-scale and diverse data. Using a combination of literature review, experimental evaluation, and comparative analysis, the research investigates how these models perform under increasing data size, varying feature complexity, and mixed data modalities. Key performance metrics, including accuracy degradation, training time escalation, memory consumption, and scalability constraints, are examined to identify critical thresholds where conventional techniques begin to fail. The results show that traditional models exhibit significant performance drops, resource saturation, and reduced robustness when faced with high-dimensional or heterogeneous datasets, particularly in comparison to modern deep learning and distributed learning approaches. These findings align with earlier theoretical studies but provide new empirical evidence that quantifies failure points and broadens the understanding of scalability limitations. The study concludes that while classical machine learning approaches remain effective for small and structured datasets, they are increasingly unsuitable for contemporary data-intensive environments. This research highlights the necessity of transitioning toward more scalable, adaptive, and representation-rich models to meet current and future data challenges.
A Probabilistic Decision Model for AI-Driven Optimization in Highly Complex Systems Riandari, Fristi; Panjaitan, Firta Sari
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 2 (2025): May: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

This research proposes a novel Probabilistic Decision Model (PDM) designed to address the challenges of optimization in highly complex systems characterized by high-dimensional states, nonlinear interactions, and deep uncertainty. Traditional deterministic, heuristic, and deep learning-based methods often fail to provide reliable decisions under such conditions due to their limited scalability, lack of uncertainty quantification, or inability to guarantee constraint satisfaction. The proposed model integrates probabilistic constraints, expectation-based objective functions, and adaptive AI-driven scenario generation to deliver a robust and flexible optimization framework. A rigorous mathematical formulation is presented, including probability space definitions, risk measures, and feasible neighborhood rules. Validation through numerical simulations demonstrates that the model maintains high feasibility, reduces worst-case risks, and remains stable even under extreme uncertainty. Case studies in smart grid optimization, logistics routing, and manufacturing scheduling further highlight significant performance improvements over classical stochastic optimization, MDP/POMDP models, and deep reinforcement learning without probabilistic modeling. The results confirm the model’s strong scalability, enhanced uncertainty modeling, and practical relevance for real-world industrial environments. This research contributes a hybrid probabilistic-AI framework that advances the reliability, resilience, and intelligence of decision-making in modern complex systems, while opening pathways for future exploration in multi-agent coordination, automated parameter tuning, and real-time adaptive optimization.
Development of a Robust–Stochastic Optimization Framework for Enhancing Stability and Efficiency in Transportation Models Sihotang, Hengki Tamando; Simbolon, Roma Sinta
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 2 (2025): May: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

This study develops a unified robust stochastic optimization framework designed to enhance the stability, efficiency, and reliability of transportation models operating under significant uncertainty. Traditional deterministic, robust-only, and stochastic-only approaches each face limitations deterministic models fail under variability, robust models tend to be overly conservative, and stochastic models struggle under extreme disruptions. To address these gaps, the proposed framework integrates worst-case uncertainty sets with probabilistic scenario modeling, enabling decisions that remain feasible under extreme conditions while maintaining optimal performance during typical operations. The methodology includes comprehensive uncertainty modeling of travel time fluctuations, demand variability, cost changes, and network disruptions; a hybrid mathematical formulation combining robust constraints with stochastic scenarios; and an efficient algorithmic structure employing enhanced decomposition techniques and scenario filtering to reduce computational complexity. Experimental results using benchmark and real-world transportation datasets show significant improvements in solution stability, travel time reliability, cost efficiency, and network resilience compared with conventional models. The hybrid framework reduces over-conservatism, lowers operational cost by up to 25%, and increases robustness under high-variability conditions, demonstrating superior performance in both normal and disrupted environments. The study advances optimization theory by offering a scalable and computationally tractable integration of two major uncertainty-handling paradigms, while contributing to transportation modeling through a practical tool capable of supporting reliable routing, scheduling, and logistics planning. Overall, this research provides a robust and adaptive optimization strategy that strengthens decision-making under uncertainty and improves the resilience of modern transportation systems.
Exploring Representation-Based Learning Techniques: Toward More Generalized and Self-Optimizing Models Prakoso Rizky A, Galih; Situmorang, Rohani
Jurnal Teknik Informatika C.I.T Medicom Vol 17 No 3 (2025): July: Intelligent Decision Support System (IDSS)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

Representation-based learning has become a foundational pillar of modern machine learning, enabling models to extract meaningful structure from complex, high-dimensional data. This study employs a mixed-method research design that integrates theoretical analysis, systematic literature review, and empirical evaluation to investigate the effectiveness of representation-based learning techniques in developing more generalized and self-optimizing machine learning models. Through an integrated review and empirical evaluation, the research investigates how different representation mechanisms influence model generalization, robustness, and adaptability across diverse data modalities. The findings show that deep, self-supervised, and contrastive representations consistently outperform traditional feature engineering, symbolic approaches, and classical statistical models, particularly in low-data and cross-domain scenarios. However, the study also identifies critical challenges including representation collapse, bias in embeddings, high computational overhead, interpretability limitations, and catastrophic forgetting that must be addressed to realize fully autonomous learning systems. In addition to synthesizing advances such as foundation models, multimodal fusion, neuro-symbolic frameworks, and efficient edge-compatible representations, this research proposes a structured framework for evaluating representation quality and outlines conceptual enhancements for self-optimizing learning systems. Overall, the study offers theoretical insights, practical evaluation tools, and forward-looking perspectives that contribute to the development of more generalized, flexible, and self-improving machine learning models capable of meeting the demands of evolving real-world applications.

Page 12 of 13 | Total Record : 130

 9   10   11   12   13 

Filter by Year

2018 2025


Reset
Filter By Issues
All Issue Vol 17 No 5 (2025): November : Intelligent Decision Support System (IDSS) Vol 17 No 4 (2025): Intelligent Decision Support System (IDSS) Vol 17 No 3 (2025): July: Intelligent Decision Support System (IDSS) Vol 17 No 2 (2025): May: Intelligent Decision Support System (IDSS) Vol 17 No 1 (2025): March: Intelligent Decision Support System (IDSS) Vol 16 No 6 (2025): January : Intelligent Decision Support System (IDSS) Vol 16 No 5 (2024): November : Intelligent Decision Support System (IDSS) Vol 16 No 4 (2024): September: Intelligent Decision Support System (IDSS) Vol 16 No 3 (2024): July: Intelligent Decision Support System (IDSS) Vol 16 No 2 (2024): May: Intelligent Decision Support System (IDSS) Vol 16 No 1 (2024): March: Intelligent Decision Support System (IDSS) Vol 15 No 6 (2024): January : Intelligent Decision Support System (IDSS) Vol 15 No 5 (2023): November : Intelligent Decision Support System (IDSS) Vol 15 No 4 (2023): September : Intelligent Decision Support System (IDSS) Vol 15 No 3 (2023): July: Intelligent Decision Support System (IDSS) Vol 15 No 2 (2023): May: Intelligent Decision Support System (IDSS) Vol 15 No 1 (2023): March: Intelligent Decision Support System (IDSS) Vol 14 No 2 (2022): September: Intelligent Decision Support System (IDSS) Vol 14 No 1 (2022): March: Intelligent Decision Support System (IDSS) Vol 13 No 2 (2021): September: Intelligent Decision Support System (IDSS) Vol 13 No 1 (2021): March: Intelligent Decision Support System (IDSS) Vol 12 No 2 (2020): September: Intelligent Decision Support System (IDSS) Vol 12 No 1 (2020): March: Intelligent Decision Support System (IDSS) Vol 11 No 2 (2019): Informatik Vol 11 No 1 (2019): Informatika Vol 10 No 2 (2018): Informatika More Issue