cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
-
Contact Email
-
Phone
-
Journal Mail Official
-
Editorial Address
-
Location
,
INDONESIA
JOURNAL OF APPLIED INFORMATICS AND COMPUTING
Published by Politeknik Negeri Batam
ISSN : -     EISSN : 25486861     DOI : 10.3087
Core Subject : Science,
Computer Science & IT
Journal of Applied Informatics and Computing (JAIC) Volume 2, Nomor 1, Juli 2018. Berisi tulisan yang diangkat dari hasil penelitian di bidang Teknologi Informatika dan Komputer Terapan dengan e-ISSN: 2548-9828. Terdapat 3 artikel yang telah ditelaah secara substansial oleh tim editorial dan reviewer.
Arjuna Subject : -
Articles 1,006 Documents
 83   84   85   86   87 
Pemodelan Migrasi Tenaga Kerja Terinduksi Modal untuk Konvergensi Regional Indonesia Dina Rahma Wita; Muhammad Subhan
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12713

Abstract

Regional inequality in Indonesia presents a persistent structural challenge characterized by capital concentration and labor agglomeration in core regions. Conventional models often fail to capture the endogenous feedback between capital accumulation, dynamic labor migration, and regional ecological limits. To address this theoretical gap, we formulate a four-dimensional non-linear ordinary differential equation (ODE) system integrating the Solow neoclassical growth framework, logistic population dynamics bounded by regional carrying capacities (κ), and Capital-Induced Labor Migration (CILM). The model is empirically calibrated using Statistics Indonesia macroeconomic data (2020-2024), classifying regions into Core and Periphery based on a strict 2.63% Gross Regional Domestic Product (GRDP) equal-distribution threshold. Analytical validity is rigorously established through mathematical proofs of Positive Invariance and Ultimate Boundedness. Numerical simulations using the Runge-Kutta 4th Order method reveal a critical phenomenon termed "deceptive convergence" a scenario where the spatial inequality ratio artificially compresses not due to genuine economic expansion in the periphery, but solely because the core region reaches its absolute ecological saturation. Furthermore, high labor mobility sensitivity is proven to act as a divergence multiplier, triggering massive labor drains from peripheral regions. Ultimately, a synthesized policy mix simulation demonstrates that aggressive physical capital diffusion combined with strict human capital retention policies is mathematically proven as the most effective strategy to transform the agglomeration curse into equitable structural convergence.
Deep Learning-Based Detection and Classification of Rice Leaf Diseases Using ResNet-50 with Augmentation and K-Fold Cross-Validation Zaqi Kurniawan; Rizka Tiaharyadini; M Saddam Ryuga Octoramdhani; Radiz Dirgantara
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12716

Abstract

Rice sustains over half of the global population; however, leaf diseases such as blight, brown spot, and leaf smut significantly reduce crop yields. Early and accurate detection is essential for supporting sustainable agriculture practices. This study proposes a ResNet-50-based deep learning model for rice leaf diseases classification using a dataset of 1,150 field images collected from Sleman, Indonesia, which was expanded to 2,000 images through augmentation techniques, including rotation, flipping, zooming, and brightness adjustment. Model performance was evaluated using both hold-out validation and 10-fold cross-validation with accuracy, precision, recall, and F1-Score metrics. The application of data augmentation improved hold-out validation accuracy from 82.4% to 88.1%. Meanwhile, 10-fold-cross-validation yielded a substantially higher average accuracy of 99.6%. This discrepancy suggests potential sensitivity to data partitioning and indicates the need for careful interpretation, as cross-validation may introduce optimistic estimates under certain conditions. Although the proposed approach demonstrates strong performance in distinguishing visually similar diseases, this study limited by the use of a single-region dataset, which may affect generalizability. Therefore, the integration of ResNet-50, augmentation, and cross-validation shows promising results for early disease detection, while further validation on more diverse datasets is required to support it application in real-world precision agriculture systems.
Comparison of Naive Bayes, Support Vector Machine, and Indobert Methods for Classifying Public Sentiment towards the MBG Program on Platform X Hanya Abriananta; Khothibul Umam; Nur Cahyo Hendro Wibowo; Maya Rini Handayani
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12721

Abstract

This study compares the effectiveness of three machine learning algorithms, namely Naive Bayes, Support Vector Machine (SVM), and IndoBERT, in classifying public sentiment towards the Free Nutritional Meal (MBG) program on Platform X. A total of 1,176 Indonesian language tweets were collected through Selenium-based web scraping from January 1 to March 31, 2025. Sentiment labeling using a lexicon-based approach with 51 positive domain-specific words and 50 negative domain-specific words, coupled with negation pattern detection, resulted in 62.8% positive tweets and 37.2% negative tweets. Preprocessing for Naive Bayes and SVM followed a six-stage workflow including stemming through PySastrawi, while IndoBERT used a minimal preprocessing approach to retain contextual information. Feature extraction applied TF-IDF with a maximum of 1,500 features and a unigram-bigram-n-gram range, with a stratified data split of 80:20. IndoBERT achieved the highest accuracy of 81.4% with a weighted F1 score of 0.81, followed by SVM at 74.2% (F1 score of 0.74) and Naive Bayes at 72.5% (F1 score of 0.72). A Wilcoxon signed-rank test on 5-fold cross-validation confirmed that the performance difference between Naive Bayes and SVM was not statistically significant (p > 0.05). These findings provide empirical evidence for policymakers to monitor public acceptance of government nutrition programs through social media analysis.
Performance Analysis of Suricata as an Intrusion Detection System (IDS) in Detecting Slowloris Attacks on Web Servers Andika Agus Slameto; Eka Marlina Kemala Sari
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12724

Abstract

Network security on web servers is a crucial element for ensuring service availability. Slowloris represents a low-rate variant of the Denial-of-Service (DoS) attack, leveraging HTTP connection handling mechanisms by submitting perpetually incomplete requests, which deplete server connection slots without necessitating high bandwidth. This study evaluated Suricata as an Intrusion Detection System (IDS) on an Apache web server through 50 controlled Slowloris attack simulations within a VirtualBox virtual environment running Ubuntu 24.04 LTS. Three performance parameters were analyzed: (1) Detection speed, quantified as the elapsed time from attack initiation to alert generation in fast.log; (2) Detection rate (True Positive Rate), determined via a confusion matrix; and (3) System resource consumption of the Suricata process (CPU, RAM, and bandwidth). Results indicated an average Detection time of 0.346 seconds (minimum 0.168 seconds, maximum 0.979 seconds), an average detection rate of 72.84% (minimum 50.00%, maximum 94.12%), a mean CPU increase of 5.22 percentage points (from 2.03% baseline to 7.25% under attack), a 6.39 MB (59.85 to 66.24 MB) increase in RAM, and a 256.72 kbps increase in bandwidth. No false positives were consistently observed across all 50 trials. Cross-parameter analysis uncovered a non-linear relationship between detection time and detection rate, with an empirically validated optimal detection-time range of 0.25 to 0.35 seconds associated with peak detection rates above 90%, and a positive correlation (r = 0.468) between attack bandwidth intensity and Detection accuracy. These findings confirmed Suricata’s effectiveness as a lightweight early-warning mechanism for Slowloris mitigation on standard-specification web servers.
Three-Tier Disaster Logistics System Integrating GIS and MILP Optimization Danny Oka Ratmana; Muhammad Syaifur Rohman; Galuh Wilujeng Saraswati; Filmada Ocky Saputra; Aprilyani Nur Safitri; Imanuel Harkespan
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12727

Abstract

Effective disaster logistics management requires rapid, data-driven decision support that bridges optimization theory and operational practice. Existing systems either rely on theoretical models without implementable software, on proprietary datasets that restrict independent reconstruction, or lack validated prototypes in the Indonesian disaster context — three gaps that persist across the disaster IS literature. This study presents a three-tier web-based disaster logistics management IS integrating GIS and MILP optimization, built exclusively on public data sources (BNPB DIBI and OpenStreetMap). Using Design Science Research (DSR) across five phases, the system employs an open-source stack: Laravel 11.x presentation layer, PostgreSQL 16/PostGIS data layer, and Python FastAPI as a dedicated MILP microservice. The MILP model, a two-phase lexicographic MILP formulation with trips-aware vehicle capacity constraints is solved using the PuLP 3.3.0 + CBC solver. Three integrated modules were developed: shelter management, warehouse inventory, and logistics coordination with GIS visualization. Functional testing achieved 100% pass rate across 85 automated test cases covering all system modules, with 246ms mean response time under 50 concurrent users. The MILP solver resolved a 20-shelter problem in 0.094 seconds (99.9% below the 120-second operational planning threshold); scalability testing confirms tractability from 10 to 50 shelters (0.011–0.111 seconds), with Priority-1 shelters consistently served under both sufficient and scarce fleet conditions. Sensitivity analysis confirms lexicographic priority objectives activate correctly under resource scarcity. Comparative evaluation against heuristic and metaheuristic approaches confirms exact MILP is appropriate for the strategic planning scope of this proof-of-concept (n ≤ 50 shelters). Expert validation via ISO 25010 yielded a weighted score of 4.21/5. Usability testing with 25 participants produced a SUS score of 74.8 (Grade B, above-average per established SUS benchmarks) with 88% task completion rate. The primary contributions are a MILP-IS microservices integration pattern with explicit API specification, a comprehensively documented public-data-only implementation framework, and a proof-of-concept that closes the implementation gap between disaster logistics optimization research and operational IS deployment.
Classification of the Effectiveness of Balur Therapy on Patients at the Malang Health Center Using the Decision Tree Algorithm Riski Puji Lestari; Mochammad Anshori; Wahyu Teja Kusuma
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12730

Abstract

This study addresses the classification of balur therapy effectiveness as a complementary treatment using a machine learning approach, aiming to develop an accurate, balanced, and transparent model to support clinical decision-making. The methodology employs the Decision Tree algorithm, data imbalance handling using Synthetic Minority Oversampling Technique, and model interpretation through Local Interpretable Model-Agnostic Explanations. The dataset consists of 520 medical records, reduced to 478 after preprocessing, including data cleaning, binning, and outlier removal. The results indicate that the model without data balancing achieved the highest specificity of 0.8276 at a 90:10 split ratio, while the application of Synthetic Minority Oversampling Technique improved sensitivity toward the minority class but reduced specificity. Key influential features include occupation, diagnosis, and therapy duration. The interpretability analysis demonstrates that the model can clearly explain feature contributions to predictions. This study concludes that integrating classification, data balancing, and explainable modeling enhances medical data analysis. The findings imply strong potential for developing objective and transparent clinical decision support systems.
Analysis of Illumination Invariant Method for Face Detection in Different Lighting Variations Ivan Chatisa; Siti Syahidatul Helma; Ibnu Surya
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12731

Abstract

Lighting quality is a crucial factor that affects the performance of camera-based face detection systems, especially in CCTV surveillance systems that operate in low or uneven lighting conditions. This study aims to analyze the performance of illumination-invariant preprocessing methods in improving the accuracy of human face detection under various lighting conditions. Three preprocessing approaches were compared, namely Histogram Equalization (HE), Gamma Correction (GC), and a hybrid method that combines both (GC+HE). The dataset used consists of 1415 human face images taken using a webcam with variations in five lighting conditions, four face directions, and three shooting distances. All images were processed using the Haar Cascade Classifier algorithm as the face detection method. Performance evaluation was conducted using accuracy, precision, recall, and confusion matrix analysis metrics. The test results showed that the hybrid method provided the best performance with a precision of 92.79%, accuracy of 87.49%, and recall of 89.61%, compared to the HE and GC methods used individually. This improvement indicates that the combination of lighting normalization and contrast enhancement can produce more stable and informative facial images for the detection process. The findings of this study indicate that the hybrid-based illumination invariant approach is very effective for application in real-time visual surveillance systems, especially in environments with limited lighting.
Bahasa Inggris Nasywa Azzah Nabila; Aviolla Terza Damaliana; Shindi Shella May Wara
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12734

Abstract

Floods are among the most frequent natural disasters in Indonesia, with thousands of events causing significant impacts on infrastructure damage and human lives. The substantial increase in the number of victims and flood-related damages in 2024 indicates that flood disaster mitigation efforts in Indonesia remain suboptimal. Consequently, a clustering-based analytical approach is required to understand patterns of flood impact across provinces. This study aims to cluster provinces in Indonesia based on flood-affected indicators using the Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) method with Bayesian Optimization to obtain optimal hyperparameters. This study comprises several stages, including data collection, data standardization, statistical test, data reduction, hyperparameter optimization, HDBSCAN algorithm, model evaluation, and analysis of clustering results. The results show that HDBSCAN with Bayesian Optimization yields a well-separated cluster structure with a DBCV value of 0.515. The clustering results consist of three primary clusters and one noise cluster. Cluster 0 (High Displacement & Inundation) consisting of 5 provinces, cluster 1 (High Fatality & Structural Damage) consisting of 4 provinces, cluster 2 (Low Impact) consisting of 21 provinces, and the noise cluster consisting of 8 provinces. These findings are intended to provide a foundation for the government to formulate targeted flood mitigation strategies tailored to the flood impact characteristics of each province.
Evaluation of Machine Learning Models for Classifying Diabetes and Hypertension Using Patient Data from Public Health Center X Febriyan Biopsa Minanda; Anandha Widya Putri Rahmadhina
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12736

Abstract

Major non-communicable diseases such as hypertension and diabetes mellitus continue to increase and require accurate and rapid early detection. This study aimed to evaluate the performance of several machine learning algorithms in classifying diabetes mellitus and hypertension using patient medical record data from a public health center in Bandar Lampung City. The dataset consisted of 662 patient records including age, gender, body mass index, blood glucose level, and blood pressure variables. The research stages included exploratory data analysis, data preprocessing, outlier handling using Interquartile Range (IQR), normalization using StandardScaler, and data splitting scenarios of 90:10, 80:20, and 70:30. A 10-fold cross validation approach was also applied to improve model validity. The algorithms tested were Random Forest, Support Vector Machine (SVM), Decision Tree, AdaBoost Classifier, and XGBoost Classifier. Evaluation results using accuracy, recall, precision, f1-score, and confusion matrix showed that Decision Tree achieved the best performance with 95% accuracy in the 80:20 scenario. The best model was implemented on the Streamlit platform and evaluated using the System Usability Scale (SUS) involving 23 respondents, obtaining an average score of 73.48 categorized as acceptable and good. The findings indicate that machine learning has the potential to support early screening and healthcare decision-making.  
AODV Routing Optimization in Wireless Mesh Networks Using SDN-Inspired Control and ETX-Driven Machine Learning Weight Adaptation Mochamad Yusril; Galura Muhammad Suranegara
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12737

Abstract

Wireless Mesh Networks (WMNs) require adaptive routing to sustain Quality of Service (QoS) under dynamic conditions, yet conventional AODV is limited by its reliance on hop count, which does not reflect actual link quality. This study proposes a hybrid framework integrating AODV, SDN-inspired centralized control, and Machine Learning (ML)-based weight optimization using the Expected Transmission Count (ETX). The model was evaluated in NS-3 (v3.45) through a four-scenario ablation study with 30 repetitions per scenario and OLSR as a baseline. A Gradient Boosting model trained on 885 samples generated routing weights based on seven QoS-related features. The results show that the proposed method significantly improves performance compared to standard AODV, with throughput increasing by 32.71%, delay decreasing by 40.19%, and routing overhead reduced by 38.92%, all statistically significant. The model achieved high predictive accuracy (R² = 0.9929) without overfitting, with ETX emerging as the most influential feature. Overall, the integration of SDN control and ML optimization enhances routing efficiency, stability, and adaptability in WMNs, offering strong potential for IoT and smart city applications.

Page 85 of 101 | Total Record : 1006

 83   84   85   86   87 

Filter by Year

2017 2026


Reset
Filter By Issues
All Issue Vol. 10 No. 4 (2026): August 2026 Vol. 10 No. 3 (2026): June 2026 Vol. 10 No. 2 (2026): April 2026 Vol. 10 No. 1 (2026): February 2026 Vol. 9 No. 6 (2025): December 2025 Vol. 9 No. 5 (2025): October 2025 Vol. 9 No. 4 (2025): August 2025 Vol. 9 No. 3 (2025): June 2025 Vol. 9 No. 2 (2025): April 2025 Vol. 9 No. 1 (2025): February 2025 Vol. 8 No. 2 (2024): December 2024 Vol. 8 No. 1 (2024): July 2024 Vol. 7 No. 2 (2023): December 2023 Vol. 7 No. 1 (2023): July 2023 Vol. 6 No. 2 (2022): December 2022 Vol 6 No 1 (2022): July 2022 Vol 5 No 2 (2021): December 2021 Vol. 5 No. 2 (2021): December 2021 Vol 5 No 1 (2021): July 2021 Vol. 5 No. 1 (2021): July 2021 Vol 4 No 2 (2020): Desember 2020 Vol. 4 No. 2 (2020): Desember 2020 Vol 4 No 1 (2020): Juli 2020 Vol 3 No 2 (2019): Desember 2019 Vol 3 No 1 (2019): Juli 2019 Vol 2 No 2 (2018): Desember 2018 Vol 2 No 1 (2018): Juli 2018 Vol. 2 No. 1 (2018): Juli 2018 Vol 1 No 2 (2017): Desember 2017 Vol 1 No 1 (2017): Juli 2017 More Issue