cover
Contact Name
Hindayati Mustafidah
Contact Email
jurnal.juita@gmail.com
Phone
+6285842817313
Journal Mail Official
jurnal.juita@gmail.com
Editorial Address
Gedung Fakultas Teknik dan Sains Universitas Muhammadiyah Purwokerto Jl. K.H. Ahmad Dahlan, Dukuh Waluh, Kembaran, Banyumas, Central Java, Indonesia
Location
Kab. banyumas,
Jawa tengah
INDONESIA
JUITA : Jurnal Informatika
ISSN : 20869398     EISSN : 25798901     DOI : 10.30595/JUITA
Core Subject : Science,
UITA: Jurnal Informatika is a science journal and informatics field application that presents articles on thoughts and research of the latest developments. JUITA is a journal peer reviewed and open access. JUITA is published by the Informatics Engineering Study Program, Universitas Muhammadiyah Purwokerto. JUITA invites researchers, lecturers, and practitioners worldwide to exchange and advance knowledge in the field of Informatics. Documents submitted must be in Ms format. Word and written according to author guideline. JUITA is published twice a year in May and November. Currently, JUITA has been indexed by Google Scholar, IPI, DOAJ, and has been accredited by SINTA rank 2 through the Decree of the Director-General of Research and Development Strengthening of the Ministry of Research, Technology and Higher Education No. 36/E/KPT/2019. JUITA is intended as a media for informatics research among academics, practitioners, and society in general. JUITA covers the following topics of informatics research: Software engineering Artificial Intelligence Data Mining Computer network Multimedia Management Information System Digital forensics Game
Articles 23 Documents
Search results for , issue "JUITA Vol. 14 Issue 1, March 2026" : 23 Documents clear
Evaluation of CNN Architectures for Kidney Stone Classification in Ultrasound Image Zuriati Zuriati; Sriyanto Sriyanto; Agiska Ria Supriyatna; Nurul Qomariyah; Dian Ayu Afifah; Zarnelly Zarnelly
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28352

Abstract

Kidney stone diagnosis requires fast and reliable evaluation, yet ultrasound interpretation still largely depends on clinical expertise. This study evaluates four Convolutional Neural Network (CNN) architectures, VGG16, ResNet50, MobileNetV2, and EfficientNetB0 for classifying kidney ultrasound images into Normal and Stone categories. Using a public dataset of 9,416 images, the models were assessed in terms of predictive performance and computational efficiency. MobileNetV2 achieved perfect classification performance, recording 100% accuracy, precision, recall, and F1-score, while maintaining the smallest parameter size (≈3.6M) and fastest training time (~44 s/epoch). VGG16 and ResNet50 also delivered near perfect accuracy (99.79% and 99.89%) with full recall for Stone cases. In contrast, EfficientNetB0 failed to generalize, yielding only 51.62% accuracy due to severe misclassification of Normal images. These results demonstrate that MobileNetV2 provides the most reliable and efficient solution for ultrasound based kidney stone classification, highlighting its strong potential for practical clinical deployment.
SmartHealth a Web-Based Platform for Adolescent Health Education with User Testing Rosmalina; Sutiyono Waluyo; Siti Solihat Holida; Khilda Nistrina; Devia Fitri Nopiani
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28372

Abstract

Adolescents in Indonesia continue to face barriers in accessing accurate, engaging, and relevant health education, while digital health literacy remains uneven. This study aims to develop and evaluate SmartHealth, a web-based adolescent health education platform, using an integrated theoretical framework combining the Technology Acceptance Model, Health Belief Model, Theory of Planned Behavior, and the Digital Health Literacy Model. The research employs a Research and Development approach that integrates the Borg and Gall model with the Waterfall software development method across four stages: need assessment, product design, system development, and limited field testing. A total of 60 high school students participated in usability testing using the System Usability Scale. The results show that SmartHealth achieved a mean usability score of 78.4, indicating good user acceptance, ease of use, and suitability for adolescent needs. The platform successfully provides accessible digital health content, interactive features, and online consultation support. This study concludes that SmartHealth is feasible, usable, and contextually relevant as a digital health education platform for Indonesian adolescents.
A Multispectral YOLOv8-Based System for Real-Time Object Detection and Distance Estimation in Blind Navigation Ema Utami; Erwin Syahrudin; Anggit Dwi Hartanto; Suwanto Raharjo
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28373

Abstract

Developing reliable real-time navigation systems for visually impaired individuals remains challenging, particularly in dynamic and low-light environments. This study proposes an integrated framework combining YOLOv8, OpenCV-based monocular distance estimation, and RGB–NIR multispectral imaging to enhance detection robustness and distance awareness. A dataset of 1,700 annotated images collected from diverse indoor and outdoor environments was used for training and evaluation using preprocessing techniques such as resizing, normalization, and data augmentation. System performance was evaluated using Precision, Recall, F1-Score, mean Average Precision (mAP), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Frames Per Second (FPS). Experimental results show that YOLOv8x achieved the best performance with an F1-Score of 0.91, mAP@50 of 0.74, MAE of 0.15 m, RMSE of 0.20 m, and a processing speed of 22 FPS. Multispectral RGB–NIR integration further improved low-light performance, increasing the F1-Score from 0.83 to 0.89 and reducing MAE from 0.28 m to 0.19 m with only a minor reduction in speed. These findings demonstrate that the proposed system provides an effective balance between accuracy and real-time performance for assistive navigation applications.
A Hybrid Case-Based Reasoning Framework Using KNN, Word2Vec, and Cosine Similarity for Employee Attrition Analysis Akhmad Arif Faisal Siregar; Ema Utami; Tika Novita Sari
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28523

Abstract

Employee attrition prediction remains a longstanding challenge in human resource analytics, as organizations increasingly depend on computational decision-support systems that are transparent, consistent, and operationally accountable. Conventional methods that rely solely on numerical attributes are restricted in their ability to accurately capture the structural and contextual relationships inherent in categorical and text-based employee descriptors. To overcome this limitation, the current study investigates a hybrid Case-Based Reasoning (CBR) retrieval framework that combines K-Nearest Neighbors (KNN) with Word2Vec embeddings derived from the dataset's limited textual attributes, specifically Department, Gender, EducationField, MaritalStatus, and OverTime. Eight experimental configurations were assessed to examine the impact of alternative similarity metrics and diverse feature representations. The optimal configuration of KNN, enhanced with Word2Vec embeddings and cosine similarity, attained an accuracy of 0.8526 and a weighted F1-score of 0.8000, thereby exceeding the performance of baseline models based solely on numerical features and those utilizing Manhattan distance. Nonetheless, the improvements in performance remained limited owing to dataset-specific limitations, such as class imbalance and the inherently superficial characteristics of the textual descriptors, which restrict the semantic richness of Word2Vec embeddings. Furthermore, the IBM attrition dataset does not encompass downsizing or termination situations, highlighting conceptual and ethical constraints when utilizing similarity-based predictions for high-stakes HR decisions. Overall, the findings indicate that hybrid similarity representations, particularly the combination of Word2Vec embeddings with cosine distance, can improve the structural expressiveness of CBR, although their predictive effectiveness is still limited by data sparsity and considerations of fairness.
Enhancing Geospatial House Price Prediction in Greater Jakarta Using XGBoost and ResNet18 Feature Fusion Hadi Santoso; Alif Biuti Anastasya
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28582

Abstract

Precise house price predictions play a vital role in shaping housing policies and informing investment decisions in urban regions such as Greater Jakarta, encompassing Jakarta, Bogor, Depok, Tangerang, and Bekasi. Most models rely exclusively on structured data, ignoring spatial and environmental factors that influence property prices. This study proposes a multimodal machine learning framework that integrates structured property data with Sentinel-2 satellite imagery within a geospatial context. The baseline dataset consists of 17 tabular variables. The ResNet-18 algorithm extracts visual environmental information from satellite imagery. The integration of both modalities through a late fusion strategy results in improved predictive performance. The baseline XGBoost achieved R² scores of 0.81 (log scale) and 0.79 (Rupiah), with an error of about 184 million Rupiah. The image-only model achieved an R² of 0.43, indicating moderate explanatory capability. Late fusion further improved performance, achieving R² values of 0.94 (log scale) and 0.93 (Rupiah), while reducing prediction error by over 40%.
Dual-Stage Bifurcation With Genetic Algorithm Extraction for Robust Anti-Antiforensic Steganalysis in Grayscale Images Amadeus Pondera Purnacandra; Yudi Prayudi
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28602

Abstract

Steganalysis plays a crucial role in digital forensics by detecting and retrieving hidden information embedded within digital media. Traditional statistical methods, such as Chi-Square and RS-Analysis, are computationally efficient but ineffective against adaptive steganography techniques that minimize detectable distortions. While recent deep learning-based approaches, particularly Convolutional Neural Networks (CNNs), have improved detection accuracy, most rely on single-stream architectures and focus solely on classification, neglecting the recovery of concealed payloads. This study proposes a dual-stage steganalysis framework that integrates a bifurcated CNN for enhanced detection with a genetic algorithm-based extraction pipeline for payload recovery. The bifurcation architecture extends GBRAS-Net by enabling parallel feature learning paths to capture diverse noise patterns, while the extraction module employs chromosome key encoding, Hilbert Curve scrambling, and LZMA compression to reconstruct hidden data. Evaluations on BOSSbase 1.01 and BOWS 2 datasets show that the proposed method achieves an average detection accuracy of 92.53%, outperforming the original GBRAS-Net (89.8%) and other CNN-based models by a statistically significant margin (p < 0.01). Furthermore, the extraction module achieves 100% payload recovery with perfect data integrity verification. The results demonstrate that integrating bifurcated feature learning with robust extraction addresses critical gaps in current steganalysis, offering a practical forensic tool for both detection and reconstruction of hidden information. This approach has significant potential for applications in law enforcement, cybersecurity, and intelligence operations.  
Evaluasi Kinerja Jaringan Sensor Nirkabel Berbasis Lora untuk Transmisi Data Bank Sampah Cerdas Mila Kusumawardani; Rabbani Yusuf Ghazali; Lis Diana Mustafa; Isa Mahfudi
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28746

Abstract

This study presents the design and performance evaluation of a LoRa-based Wireless Sensor Network (WSN) prototype for data transmission in a Smart Waste Bank system operating in a semi-urban environment. The proposed system adopts a hybrid network architecture, combining multi-hop communication among sensor nodes with a star topology for long-range transmission to a Raspberry Pi–based sink node. An ESP32 microcontroller integrated with a LoRa SX1278 module operating at 433 MHz is used to transmit waste volume and activity data with low power consumption. Unlike prior studies that predominantly focus on generic LoRa performance evaluation, this research provides a contextual and application-driven assessment of LoRa-based WSN communication tailored to community-scale Smart Waste Bank operations under non-line-of-sight conditions. System performance was evaluated using RSSI, SNR, and PDR at transmission distances ranging from 10 m to 1000 m. Experimental results show that RSSI decreases from −56 dB to −112 dB and SNR from 6.3 dB to 1.0 dB as distance increases, while the system maintains reliable communication with an average PDR of 97% up to 800 m. Distinct from existing LoRa performance studies, this work explicitly integrates a hybrid multi-hop and star WSN topology within a real-world smart waste bank deployment, providing empirical communication benchmarks that are directly applicable to community-scale IoT implementations rather than controlled or generic test scenarios.
Comparing GAN, Diffusion, and Diffusion-GAN for Single-Image Deraining of UAV Imagery Salsabilah Aulia Rahman; Laksmita Rahadianti
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.28994

Abstract

Single-image deraining for Unmanned Aerial Vehicle (UAV) imagery remains challenging due to non-uniform rain patterns, motion blur, and real-time processing requirements. Existing generative paradigms, including Generative Adversarial Networks (GAN), Diffusion, and Diffusion–GAN, each face inherent trade-offs among restoration quality, stability, and efficiency. To address the lack of unified and fair benchmarking across these paradigms, this study presents a systematic and controlled comparative evaluation of three representative models, including TBGAN, WeatherDiff, and SupResDiffGAN, to assess their relative performance in UAV deraining tasks. The models are evaluated on the UAV-Rain1K and Rain100L datasets using PSNR, SSIM, and inference efficiency metrics to support informed selection of paradigms for UAV applications. Experimental results show that WeatherDiff achieves the highest fidelity with 19.99 dB PSNR, 0.8375 SSIM on UAV-Rain1K and 29.51 dB PSNR, 0.9093 SSIM on Rain100L. TBGAN yields sharper details but lower structural consistency, whereas SupResDiffGAN offers balanced performance with 19.03 dB PSNR and 0.7053 SSIM on UAV-Rain1K and 28.51 dB PSNR and 0.8681 SSIM on Rain100L, with faster inference. These findings highlight the practical trade-offs among the three paradigms and demonstrate that diffusion–GAN frameworks provide the most practical solution for UAV deraining, combining diffusion stability with adversarial sharpness for real-time restoration.
Peramalan Penjualan Obat Pasca Masa Paten: Studi Perbandingan Model Statistik dan Pembelajaran Mesin Sebastian Miguel; Sfenrianto Sfenrianto
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.29008

Abstract

Volatility in the pharmaceutical industry can be caused by expiration of drug patents, leading to a gap between actual and target sales values, which necessitates accurate sales forecasting for pharmaceutical marketers. This study utilizes the sales data from PT. Q, an Indonesian pharmaceutical firm. The comparative performance within the specific context of the post-patent period for pharmaceutical sales remains relatively unexplored. This research aims to compare forecasting models for post-patent pharmaceutical sales. The research method utilized is based on the CRISP-DM data mining framework. The forecasting process is done on a 4.5-year timeframe using forecasting models such as ARIMA, SARIMA, LSTM, and Prophet. The results show that multivariate LSTM works better for forecasting in smaller aggregations in the dataset such as by product type and branch, with a R² score value of up to 0.64 in the aggregation level of Bandung_Sales, and with the smallest error metric values, such as MAE in many aggregation levels, example being regional sales, such as Lampung_Sales with 1.31 and Makassar_Sales with 0.26, which outperforms the other compared models in the majority of cases. This research concludes that multivariate LSTM is a better way to replace outdated methods to set sales targets.
Performance Evaluation of Tuned and Untuned Machine Learning Models in Speech Emotion Recognition Muhammad Hudzaifah Nasrullah; Dede Cahyadi; Tilly Raycitra Widya; Ewin Suciana; Lilik Tiara Giantri
JUITA: Jurnal Informatika JUITA Vol. 14 Issue 1, March 2026
Publisher : Department of Informatics Engineering, Universitas Muhammadiyah Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30595/juita.v14i1.29015

Abstract

This analysis takes on a comparative review of three distinct machine learning approaches: Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), and Random Forest (RF) to ascertain emotional states in verbal communication by utilizing the RAVDESS resource. In this review, we perform a strategy that unites audio feature extraction, model training with or without tweaks to hyperparameters, and evaluation via metrics including accuracy, precision, recall, and F1-score. The assessment shows that, before any refinement, SVM secured the utmost accuracy of 79%, trailed by MLP at 76% and RF at 71%. Following optimization, only SVM exhibited an enhancement, reaching 80%, whereas MLP and RF displayed negligible or no improvement. An examination of the confusion matrix revealed that SVM produced the most uniformly distributed predictions and effectively reduced misclassification errors, particularly within the emotion categories of “calm” and “happy.” This investigation offers empirical substantiation of SVM as a robust baseline model for speech emotion recognition in localized settings, while simultaneously providing insights into model optimization and development that could inform future implementations in speech-based human–computer interaction.

Page 2 of 3 | Total Record : 23