cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Budi Hermawan
Contact Email
-
Phone
+62081703408296
Journal Mail Official
info@kdi.or.id
Editorial Address
Jl. Flamboyan 2 Blok B3 No. 26 Griya Sangiang Mas - Tangerang 15132
Location
Kab. tangerang,
Banten
INDONESIA
bit-Tech
Published by Komunitas Dosen Indonesia
ISSN : 2622271X     EISSN : 26222728     DOI : https://doi.org/10.32877/bt
Core Subject : Science,
Computer Science & IT
The bit-Tech journal was developed with the aim of accommodating the scientific work of Lecturers and Students, both the results of scientific papers and research in the form of literature study results. It is hoped that this journal will increase the knowledge and exchange of scientific information, especially scientific papers and research that will be useful as a reference for the progress of the State together.
Arjuna Subject : Ilmu Komputer (semua kategori) - Ilmu Komputer (Lain-Lain)
Articles 642 Documents
 44   45   46   47   48 
Comparison of Fine-Tuning InceptionV3 and Xception for Eye Disease Classification Based on Fundus Images Irsyad Rafi Naufaldi; Ani Dijah Rahajoe; Eva Yulia Puspaningrum
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3195

Abstract

Eye diseases represent a major global health concern that can lead to visual impairment and even blindness if not detected early. The shortage of ophthalmologists and unequal distribution of medical services highlight the need for automatic eye disease detection system increasingly essential. Therefore, the role of Artificial Intelligence (AI), particularly Deep Learning, is highly needed. This study aims to compare the performance of two CNN architectures InceptionV3 and Xception. Unlike previous studies, this paper provides a comparative Fine-Tuning analysis of two CNN models on multiclass eye disease. The approach applied is transfer learning with a fine-tuning technique on several final layers to achieve higher accuracy by optimizing pretrained models using large-scale datasets such as ImageNet. The dataset consists of 4,184 fundus images covering multiple eye disease with balanced class distribution, ensuring diversity that supports model generalization. Divided into train, valid, and test sets with a ratio of 70:15:15. The training employed Adam optimizer, a batch size of 16, a learning rate of 0.0001, and implements early stopping to prevent overfitting. The performance of the model was assessed using evaluation metrics including accuracy, precision, recall, and F1-score. Experimental results indicate that the Xception model achieved superior performance with an accuracy of 87.78%, precision of 0.89, recall of 0.88, and an F1-score of 0.88, outperforming InceptionV3 with an accuracy of 85.56%, indicates the model is reliable for preliminary diagnosis. These findings suggest that the architecture in Xception is more efficient in extracting features from limited yet complex medical datasets.
DR-ARMA Method for Predicting Seblak Safety Stock Indri Sari Dewi; Jimmy Tjen
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3196

Abstract

The rapid growth of the small culinary sector intensifies competition and operational complexity, particularly in managing inventory levels to ensure product availability and customer satisfaction. This study focuses on Warung Seblak Uwak, which uses a dual demand structure: a customizable prasmanan model and predetermined Seblak bundled packages. This research specifically analyzes the demand for these bundled packages, which, despite being standardized, still exhibit complex and volatile daily patterns influenced by overall store traffic. Accurate stock management for these items is crucial for maintaining profit margins and minimizing ingredient spoilage. To address the challenge of this unpredictable demand and optimize inventory in the small F&B context, this study pioneers the application of the Demand Response-AutoRegressive Moving Average (DR-ARMA) model. This sophisticated time-series methodology, previously confined to industrial or financial risk assessment, is novel in its capacity to adapt its forecast to recent sales anomalies in a dynamic culinary setting, offering superior predictive performance over standard methods. This application fills a critical gap in F&B forecasting literature. The research analyzes inventory risks and determines the optimal safety stock for the bundled packages using DR-ARMA (1,3). The methodology utilized 127 days of sales transaction records from Warung Seblak Uwak, followed by rigorous testing. The model achieved a RMSE of 5.9310, demonstrating high predictive acacuracy. The resulting safety stock recommendations offer a quantified and robust strategy for micro and small culinary enterprises, specifically concerning their standardized products, to significantly mitigate stockout risks and reduce waste, thereby improving operational efficiency and profitability.
Classification Tuberculosis on Chest X-Ray Images Using Backpropagation Neural Network Ananda Ayu Puspitaningrum; Anggraini Puspita Sari; Muhammad Muharrom Al Haromainy
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3197

Abstract

Tuberculosis is an infectious disease that primarily affects the lungs and remains a major health concern due to the difficulty of diagnosis through manual interpretation of chest X-ray images. This study aims to develop an automatic tuberculosis classification system using the Backpropagation Neural Network (BPNN) method to improve diagnostic accuracy. The dataset used in this study was obtained from the Kaggle Tuberculosis (TB) Chest X-ray Dataset, consisting of 7.000 images divided into two classes normal and tuberculosis. The research stages include image preprocessing such as conversion to grayscale, resizing to 256×256 pixels, contrast enhancement using histogram equalization, and noise reduction using a median filter. Experiments were conducted by varying the number of hidden layers 2, 3, and 4 to analyze the effect of network architecture complexity on classification performance. The results showed that the configuration with 2 hidden layers and [100 50] neurons achieved the best performance with an accuracy of 93.57%. The findings indicate that deeper network architectures do not always guarantee higher accuracy and may increase computational load. Overall, this configuration provides an optimal balance between learning capability and accuracy, demonstrating the potential of the BPNN method in supporting early and reliable tuberculosis detection through machine learning based chest X-ray image analysis for clinical decision support.
Comparative Analysis of Performance Aspects Between Chroma and Pgvector as a Vector Database Ali Zaenal Abidin; Mychael Maoeretz Engel
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3198

Abstract

System architects face a critical choice between specialized vector databases like ChromaDB and general-purpose options like PostgreSQL with the PGVector extension. This decision profoundly impacts TCO and system viability, yet holistic performance data under resource constraints is scarce. We answer whether a specialized or generalized architecture provides superior operational efficiency and accuracy when resources are limited and providing an evidence based guide for navigating the trade offs between cost, speed, and accuracy. We conducted 119 tests on the Deep1M dataset within a resource-constrained 4GB RAM Docker, measuring latency, ingestion speed, storage overhead, and recall accuracy. The results reveal a stark architectural trade-off.  ChromaDB delivers highly consistent, low query latency, with only a 1.3-fold performance degradation as data scales. However, this speed comes with significant operational costs:-massive storage inefficiency averaging 395 times the raw data size and severe ingestion bottlenecks, showing a 491.7 fold slowdown. Conversely, PostgreSQL with PGVector demonstrates resource efficiency. Its storage overhead is minimal at 3-4 times the raw data size, and it provides 7.0 times better ingestion scalability. Crucially, it achieves statistically superior accuracy at production scale (≥250K vectors), delivering near-perfect 99.6-99.8% recall compared to ChromaDB's 91-95%. The trade-off is performance variability, where poorly tuned PostgreSQL queries can be up to 16.6 times slower than ChromaDB. We conclude that for dynamic production applications where TCO, scalability, and high accuracy are priorities, PGVector is more viable. ChromaDB's predictable latency is better suited for latency-critical applications with static data, but only if its high operational costs are acceptable.
Comparative Analysis of YOLOv8 and YOLOv11 in Breast Lesion Detection Bridget Beatrix Claire; Daniel Martomanggolo Wonohadidjojo
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3202

Abstract

Breast cancer remains one of the leading causes of mortality among women worldwide, emphasizing the need for accurate and efficient diagnostic tools. Ultrasound imaging is widely used for breast lesion screening due to its affordability and safety, yet manual interpretation often suffers from variability and subjectivity. Recent advancements in deep learning, particularly the YOLO (You Only Look Once) family, have demonstrated strong potential for real-time medical image detection and segmentation. This study aims to compare the performance of YOLOv8m-seg and YOLOv11m-seg models in detecting and segmenting breast lesions from ultrasound images to determine which model offers a better balance between accuracy, sensitivity, and computational efficiency. Two public ultrasound datasets were employed to ensure data diversity and evaluation fairness. Both models were trained under identical preprocessing, augmentation, and hyperparameter settings using 640×640 input resolution and the AdamW optimizer. Model performance was evaluated through Precision, Recall, F1-score, mAP@0.5, mAP@0.5:0.95, Mask Precision, and Inference Time metrics. The experimental results show that YOLOv11m-seg outperformed YOLOv8m-seg in precision (0.859), mask accuracy (0.859), and inference time (16.7 ms), while YOLOv8m-seg maintained slightly higher recall (0.736). YOLOv11m-seg demonstrated stronger generalization across heterogeneous datasets and superior boundary segmentation. YOLOv11m-seg achieved the best overall performance and is more suitable for real-time clinical applications. This study contributes empirical benchmarks for future Computer-Aided Diagnosis (CAD) development and highlights the potential of modern YOLO architectures in improving breast ultrasound lesion detection accuracy and efficiency.
Classification of Jombang Batik Motifs Using Ensemble Convolutional Neural Network Riza Satria Putra; Muhammad Muharrom Al Haromainy; Achmad Junaidi
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3204

Abstract

Batik, recognized by UNESCO as an Intangible Cultural Heritage, presents complex visual patterns that challenge automated classification systems. The intricate variations in texture, color, and geometry across motifs often lead to inconsistent performance in single Convolutional Neural Network (CNN) models, which struggle to generalize across subtle inter-class differences. To address these limitations, this study implements an Ensemble CNN framework to classify six Ploso Jombang batik motifs Garudan, Merak Kinasih Keyna Galeri, Ploso Bersemi, Jombang Berseri, Sulur Kangkung, and Burung Hong from a dataset of 2,134 images. The proposed approach integrates three pre-trained architectures EfficientNetB0, ResNet18, and VGG16 through a stacking ensemble strategy to leverage complementary feature extraction capabilities. Experimental results demonstrate that EfficientNetB0 achieved the highest individual accuracy (94%), while VGG16 recorded the lowest (60%). When combined, the ensemble configurations EfficientNetB0 + VGG16 and EfficientNetB0 + ResNet18 achieved peak test accuracies of approximately 96.88% on 321 test samples, reflecting a 2.88% improvement over the best single model. Confusion Matrix analysis confirmed robust model stability, with 100% accuracy for motifs such as Ploso Bersemi and Sulur Kangkung. These results validate that ensemble learning effectively mitigates overfitting and enhances generalization by aggregating diverse visual representations. The proposed model thus provides a reliable computational framework for automated batik classification and digital cultural preservation, supporting Indonesia’s efforts to document, catalog, and sustain its traditional heritage through artificial intelligence–driven methods.
Prediction Of Student Stress Levels Based on Random Forest and The Dass-21 Questionnaire Raisa Indira Zahra; Anna Dina Kalifia
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3208

Abstract

Academic stress is a psychological condition commonly experienced by students due to increasing academic, social, and emotional demands. Without early detection, this issue can negatively impact mental health and academic performance. This study aims to develop and evaluate a machine learning–based model using the Random Forest algorithm to predict students’ stress levels based on the Indonesian version of the Depression Anxiety Stress Scale-21 (DASS-21). Data were collected from 143 university students in Yogyakarta who completed the DASS-21 questionnaire, and stress subscale scores derived from seven items were multiplied by two and categorized into five levels: Normal, Mild, Moderate, Severe, and Extremely Severe. The dataset was then cleaned, labeled, normalized, and split into training and testing subsets (60:40) using stratified sampling. Model performance was evaluated using accuracy, macro-F1, Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). The Random Forest model achieved an accuracy of 87.93%, macro-F1 of 0.7047, MAE of 0.121, and RMSE of 0.347, with the best performance observed in the Severe (F1 = 0.9387) and Normal (F1 = 0.9230) categories. To enhance practical usability, the model was deployed in a web-based system named StressPredict, which provides real-time predictions, class probabilities, and an analytical dashboard for monitoring student populations. The findings confirm that the Random Forest algorithm is effective for multi-level stress classification and demonstrates strong potential as a digital mental health monitoring tool in higher education environments, supporting early screening and informed interventions for student well-being.
Waste Classification Using YOLOv8 and One Factor At a Time Muhammad Aldi Maulana; Eva Yulia Puspaningrum; Ani Dijah Rahajoe
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3209

Abstract

Solid waste management has become a significant global environmental challenge that affects both ecosystem sustainability and human well-being. The increasing volume of waste generated from daily human activities highlights the urgent need for technology-based solutions that support efficient waste sorting, recycling, and resource recovery. This study proposes an automatic waste classification system using the YOLOv8 algorithm, a state-of-the-art deep learning model capable of performing real-time object detection with high accuracy. A dataset consisting of 1,800 labeled waste images representing five main categories plastic, glass, metal, paper, and organic was used for model training and evaluation. To enhance performance, the One Factor at a Time (OFAT) approach was applied for hyperparameter optimization, focusing on learning rate, batch size, and number of epochs. Two models were compared: the default YOLOv8 configuration and the optimized YOLOv8 OFAT model. Experimental results show that the optimized YOLOv8 OFAT achieved a mAP@0.5:0.95 of 86.1%, slightly higher than the default YOLOv8 model with 85.8%. Although the improvement of 0.3% appears modest, it indicates better model consistency and reliability across various data conditions. The integration of the OFAT technique into YOLOv8 represents a novel contribution, demonstrating that systematic hyperparameter tuning can significantly enhance the efficiency and robustness of automated waste detection systems, thereby supporting environmental sustainability and the realization of a green economy.
Implementation of GRU with Attention Mechanism for Classifying Lung Diseases from Respiratory Sounds Kartika Sari; Anggraini Puspita Sari; Afina Lina Nurlaili
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3210

Abstract

Early and accurate detection of lung diseases plays a crucial role in improving treatment outcomes and reducing mortality rates, particularly in low-resource healthcare settings. Conventional auscultation using a stethoscope is a fundamental, fast, and affordable method for initial lung examination. However, its effectiveness is limited by subjectivity, as it depends on the examiner’s expertise and can be influenced by environmental noise. To overcome these limitations, this study proposes a deep learning approach for lung diseases classification using a combination of Gated Recurrent Unit (GRU) and Attention Mechanism with log Mel spectrogram as an input based on respiratory sound. Unlike previous works that employed standalone methods such as GRU or CNN, the integration of Attention mechanism in this study allows the model to focus on prominent temporal patterns within respiratory sounds, thereby enhancing classification accuracy. Experiments were conducted on the ICBHI 2017 dataset, which underwent preprocessing stages consisting of minor class removal, recording location restriction, data augmentation, and log Mel spectrogram feature extraction. The test results show that the model produces high performances with an accuracy of 90.85%, precision of 93%, recall of 90.85%, and an F1-score of 91.14%, outperforming several works that reported in prior studies. These results demonstrate the effectiveness of combining GRU and Attention mechanism in capturing the temporal features of respiratory signals. Future research could focus on enhancing model robustness through improved data quality, other model architecture, and multimodal integration for broader clinical applicability.
Space-Time Modeling for Forecasting Large Red Chili Prices Based on Significant Parameter Selection Sandria Amelia Putri; Mohammad Idhom; Aviolla Terza Damaliana
bit-Tech Vol. 8 No. 2 (2025): bit-Tech
Publisher : Komunitas Dosen Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32877/bt.v8i2.3212

Abstract

FVolatile fluctuations in large red chili prices pose a persistent challenge to Indonesia’s food security and regional economic stability, as price shocks directly affect household purchasing power, inflation, and agricultural income. Addressing this issue requires a forecasting framework that captures both spatial interdependence among producing and consuming regions and temporal price dynamics. This study develops an advanced forecasting model for large red chili prices in East Java covering Malang Regency, Banyuwangi Regency, and Surabaya City using the Generalized Space-Time Autoregressive–Seemingly Unrelated Regression (GSTAR-SUR) method. The model integrates the Generalized Least Squares (GLS) approach to enhance parameter estimation efficiency under correlated residuals and applies a partial t-test–based parameter elimination procedure to retain only statistically significant predictors. Compared to traditional univariate time-series approaches such as ARIMA, GSTAR-SUR more effectively captures cross-regional price linkages and residual dependencies, yielding higher forecasting accuracy. The best-performing specification, GSTAR-SUR(3,1)-I(1) with a uniform spatial weighting matrix, achieved RMSE = 1426.73, MAPE = 3.29%, and R² = 0.8482, representing a substantial improvement in precision over conventional GSTAR and ARIMA models. Fourteen-day forecasts reveal region-specific dynamics: a mild downward trend in Malang, an initial rise followed by decline in Banyuwangi, and relative stability in Surabaya. These results demonstrate that the GSTAR-SUR framework can effectively model complex spatio-temporal dependencies in commodity markets and serves as a practical decision-support tool for policymakers in stabilizing food prices, improving distribution strategies, and strengthening agricultural market resilience across East Java.

Page 46 of 65 | Total Record : 642

 44   45   46   47   48 

Filter by Year

2018 2026


Reset
Filter By Issues
All Issue Vol. 8 No. 3 (2026): bit-Tech Vol. 8 No. 2 (2025): bit-Tech Vol. 8 No. 1 (2025): bit-Tech Vol. 7 No. 3 (2025): bit-Tech Vol. 7 No. 2 (2024): bit-Tech Vol. 7 No. 1 (2024): bit-Tech Vol. 6 No. 3 (2024): bit-Tech Vol. 6 No. 2 (2023): bit-Tech Vol. 6 No. 1 (2023): bit-Tech Vol. 5 No. 3 (2023): bit-Tech Vol. 5 No. 2 (2022): bit-Tech Vol. 5 No. 1 (2022): bit-Tech Vol. 4 No. 3 (2022): bit-Tech Vol. 4 No. 2 (2021): bit-Tech Vol. 4 No. 1 (2021): bit-Tech Vol. 3 No. 3 (2021): Remote Delivery Vol. 3 No. 2 (2020): Datamining Vol. 3 No. 1 (2020): Distance Learning Vol. 2 No. 3 (2020): Pandemik ICT Vol. 1 No. 3 (2019): Learning Synchronous and Asynchronous in the Industry 4.0 Vol. 2 No. 2 (2019): Support System Vol. 2 No. 1 (2019): Data Mining and Green Technology Vol. 1 No. 2 (2018): Data and Information Quality Vol. 1 No. 1 (2018): First Impression on Computing More Issue