cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
Husni Teja Sukmana
Contact Email
husni@bright-journal.org
Phone
+62895422720524
Journal Mail Official
jads@bright-journal.org
Editorial Address
Gedung FST UIN Jakarta, Jl. Lkr. Kampus UIN, Cemp. Putih, Kec. Ciputat Tim., Kota Tangerang Selatan, Banten 15412
Location
Kota adm. jakarta pusat,
Dki jakarta
INDONESIA
Journal of Applied Data Sciences
Published by Bright Publisher
ISSN : -     EISSN : 27236471     DOI : doi.org/10.47738/jads
Core Subject : Science, Engineering,
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management
One of the current hot topics in science is data: how can datasets be used in scientific and scholarly research in a more reliable, citable and accountable way? Data is of paramount importance to scientific progress, yet most research data remains private. Enhancing the transparency of the processes applied to collect, treat and analyze data will help to render scientific research results reproducible and thus more accountable. The datasets itself should also be accessible to other researchers, so that research publications, dataset descriptions, and the actual datasets can be linked. The journal Data provides a forum to publish methodical papers on processes applied to data collection, treatment and analysis, as well as for data descriptors publishing descriptions of a linked dataset.
Arjuna Subject : Ilmu Komputer (semua kategori) - Kecerdasan Buatan
Articles 583 Documents
 55   56   57   58   59 
Monet2Photo: Reverse Style Transfer using CycleGAN with Impressionism-to-Reality Domain Wijaya, George Kerry; Chang, Shining Sunny; Saputra, Jonathan; Chloe, Annabelle; Johan, Monika Evelin
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1222

Abstract

The application of artificial intelligence in artistic image transformation has primarily focused on converting real-world photographs into stylized artworks. In contrast, the inverse task of reconstructing photorealistic images from paintings remains relatively underexplored and presents substantial technical challenges. This study aims to investigate the feasibility and limitations of reverse style transfer by translating impressionist paintings into realistic photographic images, using the works of Claude Monet as a representative case. The main contribution of this research lies in providing a critical examination of reverse image translation under extreme domain gaps, rather than proposing aesthetic enhancement. An unpaired image-to-image translation framework based on CycleGAN is employed to learn mappings between painting and photographic domains without relying on paired data. The methodology is conceptually grounded in adversarial learning combined with cycle consistency constraints to encourage structural preservation while attempting to reconstruct plausible visual features. The experimental setup utilizes a dataset consisting of 300 Monet paintings and 7,028 real photographs, with targeted data augmentation applied to the painting domain to address data imbalance. Prior to model training, exploratory data analysis is conducted to characterize domain discrepancies through visual and statistical comparisons, including color distribution analysis, grayscale intensity patterns, texture descriptors, and dimensionality reduction. Model performance is evaluated through controlled experiments using distribution-based distance measures and qualitative visual inspection. The results indicate that while the model is capable of preserving coarse spatial layouts and generating diverse outputs without memorization, it struggles to recover high-fidelity textures, illumination, and contrast required for photorealistic reconstruction. These findings highlight the inherent limitations of classical CycleGAN architectures for reverse style transfer and suggest the need for more expressive models and stronger constraints in future research on art-to-reality image translation.
Cognitive and Technological Factors Shaping Students’ Sustained Use of ChatGPT in Higher Education Aribowo, Arnold; Hery, Hery; Widjaja, Andree Emmanuel
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1291

Abstract

This study examines the cognitive and technological factors shaping students' sustained use of ChatGPT in Indonesian higher education. Despite the rapid adoption of generative Artificial Intelligence (AI) in education, a clear understanding of the factors sustaining continued engagement with such systems remains limited. While continuance intention has been widely examined, the application of the Expectation–Confirmation Model (ECM) in generative AI contexts remains underexplored. This gap is especially evident when considering the role of AI-specific system attributes in shaping post-adoption evaluations. Although ECM has been extended with various constructs in prior studies, the specific integration of AI characteristics, particularly perceived intelligence and anthropomorphism, has not been explored in generative AI use in education, especially within Indonesian higher education. To address this gap, a multi-theoretic framework integrating ECM and AI characteristics was developed. Data from 322 Indonesian students were analyzed using Partial Least Squares-Structural Equation Modeling. All ten hypotheses were supported, and the model explains 43.3% of the variance in continuance intention (R² = 0.433). Perceived Intelligence strongly influences Perceived Anthropomorphism with a path coefficient of 0.591, representing the strongest relationship in the model, while other paths demonstrate moderate or modest effects. The findings confirm ECM's robustness in generative AI settings and highlight the pivotal role of AI characteristics in shaping post-adoption evaluations and sustained use. These results contribute to the growing body of research on generative AI adoption in education by demonstrating how system intelligence and human-like interaction jointly influence continuance intention. The findings also offer practical guidance for AI developers to enhance system intelligence and natural interaction. Future research could explore how students experience AI over time and what shapes their sustained use using different research methods.
Adaptive k-Nearest Neighbor Learning for Robust Modal Regression on Multimodal and Heavy-Tailed Data Sutarman, Sutarman; Herawati, Netti; Nababan, Adli Abdillah
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1221

Abstract

Modal regression has attracted increasing attention as an alternative to mean-based regression, particularly in settings characterized by heteroscedasticity, multimodal conditional distributions, and heavy-tailed noise. In such scenarios, estimators based on central tendency may yield predictions that fall in low-density regions of the response space. This paper proposes an adaptive k-nearest neighbor framework for modal regression that integrates entropy-guided neighborhood selection with nonparametric mode estimation, including MeanShift clustering and one-dimensional kernel density estimation. The proposed approach adjusts neighborhood size based on local uncertainty, allowing the regression model to adapt to variations in data density without relying on a globally fixed parameter. Extensive experiments on simulated datasets and real-world benchmarks demonstrate that adaptive modal regression methods generally reduce or stabilize prediction errors relative to fixed-k modal regression and classical kNN mean and median estimators, particularly under heteroscedastic and multimodal conditions, although the magnitude of improvement varies across scenarios. Statistical tests confirm significant differences in most experimental settings, with practical gains ranging from incremental to substantial depending on data complexity. In addition to accuracy, computational behavior is explicitly examined. The findings show a trade-off between computational cost and predictive robustness: entropy-guided adaptive modal regression requires additional runtime due to neighborhood adaptation and density estimation, but this overhead increases proportionally with sample size and remains manageable for medium-sized datasets. Based on these results, adaptive modal regression provides a useful and flexible alternative for regression tasks involving complex and heterogeneous data distributions where robustness is prioritized over minimal computation time.
Optimized Hybrid CNN-LSTM Model for Predicting Transportation Sector Stock Prices Using Optimizer and Activation Function Tuning Gunawan, Fiena; Kristiyanti, Dinar Ajeng
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1323

Abstract

Stock price prediction is inherently complex due to nonlinear dynamics and high volatility, particularly in Indonesia’s transportation sector, which experienced significant inflationary pressure and extreme instability during and after the COVID-19 pandemic. These disruptions introduced structural breaks and regime shifts, intensifying non-stationary market behavior and increasing forecasting uncertainty. Such conditions create an urgent need for robust predictive information systems capable of supporting investment decision-making and risk management in highly volatile environments. However, standalone recurrent models such as Long Short-Term Memory (LSTM) often struggle to capture local micro-patterns and long-term dependencies. Moreover, prior studies have rarely implemented systematic hyperparameter optimization, resulting in inconsistent predictive performance across stocks with heterogeneous volatility. In contrast, Convolutional Neural Networks (CNN) extract local patterns and short-term nonlinear features, making them effective for modeling high-frequency fluctuations. This study proposes a systematically optimized hybrid CNN-LSTM model to forecast transportation sector stock prices using daily OHLC data from 2020-2025. The research framework follows the Cross Industry Standard Process for Data Mining (CRISP-DM) methodology, encompassing business understanding, data understanding, data preparation, modeling, evaluation, and deployment. Prior to modeling, preprocessing includes data cleaning, Min-Max normalization, and sliding window transformation to construct supervised learning sequences. CNN is employed to extract localized nonlinear features and reduce noise, while LSTM models long-term temporal dependencies. Model performance is evaluated using MAE, MSE, RMSE, MAPE, and R². Results show that the optimized CNN-LSTM model outperforms the baseline across all stocks. The highest R² of 0.9725 is obtained from one stock, indicating strong performance. In addition, the average R² improves from 0.8736 to 0.9483, an increase of 0.0747 (8.55%). The best results are achieved using ReLU with Adam and Nadam optimizers, demonstrating improved convergence and generalization. These findings highlight the effectiveness of optimized hybrid deep learning models for forecasting in nonlinear and non-stationary financial markets.
Optimizing Monkeypox Detection Using Advanced Class Imbalance Handling Methods: Smote, Smote-Enn, Smote-Tomek, Borderline-Smote Rizki, Fahlul; Widowati, Widowati; Widodo, Catur Edi
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1286

Abstract

Monkeypox is a zoonotic viral disease with increasing global concern due to its rapid spread and potential public health impact. Accurate and timely detection is crucial, yet the development of machine learning-based detection systems is often challenged by class imbalance in clinical datasets, leading to biased predictions towards majority classes. This study systematically evaluates the effectiveness of various class imbalance handling techniques, including SMOTE, Borderline-SMOTE, SMOTE-ENN, and SMOTE-Tomek, on the performance of ensemble learning algorithms, specifically Random Forest and Gradient Boosting, for monkeypox detection. Using a dataset of 25,000 synthetic patient records with 11 clinical features, models were trained and validated through stratified 5-fold cross-validation. Performance metrics including accuracy, precision, recall, F1-score, and Area Under the Curve (AUC), along with ROC analysis, were employed to assess the impact of each augmentation method. Results indicate that hybrid methods, particularly SMOTE-ENN, significantly improve recall and F1-score, improving the detection of clinically important monkeypox-positive cases while maintaining adequate discriminative ability. Standard SMOTE and SMOTE-Tomek provide stable performance across metrics, whereas Borderline-SMOTE shows lower recall despite high precision. These findings highlight the importance of selecting appropriate class imbalance handling strategies tailored to the clinical objective, emphasizing sensitivity in detecting positive monkeypox cases. The study provides practical guidance for implementing reliable and robust machine learning models in early monkeypox detection, contributing to improved clinical decision-making and public health interventions.
Artificial Intelligence, Transformational Leadership, and Job Performance: Mediating Role of Job Engagement and Moderating Role of Work Passion Nguyen, Phuong Thao; Nguyen, Phuc Quy Thanh; Nguyen, Minh Tuan
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1266

Abstract

This study investigates the relationships among Artificial Intelligence (AI), perceived usefulness of AI, transformational leadership, job engagement, and job performance, with the moderating role of work passion. Drawing on the Job Demands–Resources (JD–R) model and the Technology Acceptance Model (TAM), the study proposes a research model explaining how technological and leadership resources jointly influence employee performance in the context of digital transformation. A quantitative approach was employed, with data collected through an online survey of 345 employees at five leading joint-stock commercial banks in Vietnam. Partial Least Squares Structural Equation Modeling (PLS-SEM) was applied to test the proposed hypotheses. The findings reveal that perceived usefulness of AI is the strongest indirect predictor of job performance through the mediating role of job engagement. The results also confirm that transformational leadership significantly enhances employee engagement, particularly through inspirational motivation and individualized consideration. Artificial Intelligence, as an organizational resource, further strengthens engagement by reducing workload and supporting decision-making processes. Furthermore, work passion plays a moderating role in the relationship between job engagement and job performance, with harmonious passion amplifying this relationship while obsessive passion may reduce its marginal effect. These findings highlight the importance of integrating AI applications with effective leadership practices to foster employee engagement and improve job performance in modern digital organizations.
Automatic Analysis of Political Discourse: A Comparative Study of Multilingual and Large Language Models Sairanbekova, Ayaulym; Nazyrova, Aizhan; Bekmanova, Gulmira; Zhetkenbay, Lena; Yergesh, Banu; Lamasheva, Zhanar
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1118

Abstract

This paper proposes the growing importance of automated analysis of political discourse in low-resource languages, using the Kazakh language as a case study. As political communication in Kazakhstan has increasingly moved online between 2019 and 2023, the need for accurate tools to evaluate political sentiment has grown. However, limited linguistic resources in Kazakh have hindered tool development. This paper introduces the first annotated corpus of political discourse in Kazakh, comprising 3,022 sentences selected from official statements, televised debates, policy documents, and social media publications. Each text was manually annotated for political sentiment by expert linguists and political scientists, with inter-annotator agreement measured to confirm reliability. Two main methodological approaches were employed for automatic sentiment classification: adapting multilingual neural network models to the Kazakh corpus and testing advanced generative language models in scenarios with minimal training examples. Performance was evaluated using standard classification procedures. The inclusion of pragmatic features such as code-switching, rhetorical emphasis, and discursive context led to notable improvements in classification accuracy. Experimental results demonstrate that models adapted to multilingual input achieved high classification quality, with fine-tuned multilingual transformer models reaching F₁-scores of up to 0.90, while large language models reached an F₁-score of 0.94 in few-shot settings. Explicit modeling of code-switching and pragmatic features yielded an improvement of approximately 4 percentage points in F₁. This research contributes a practical resource and a methodological framework for analyzing political sentiment in underrepresented languages, highlighting the feasibility of developing high-quality automated tools for political text analysis without extensive training data.
Application of Random Data Splitting Technique for Time Series Modeling Using Feedforward Neural Network and Support Vector Regression Models Basnayake, B. R. P. M.; Chandrasekara, N. V.
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1216

Abstract

In predictive modeling, data partitioning mainly involves dividing the dataset into a training set for learning model parameters and a testing set for assessing generalizability through predictive performance. In time series forecasting, non-random data splitting is commonly used where initial observations are used for training, a subsequent portion for validation and latest observations are utilized for testing. However, the effectiveness of random data partitioning where observations are randomly distributed into training, validation and testing subsets remains underexplored within the context of time series data. This study investigates the suitability of random data partitioning in time series forecasting using both stationary and non-stationary simulated datasets, as well as real-world data. Feedforward neural network (FFNN) and Support vector regression (SVR) models were implemented, with model performance optimized through systematic trial-and-error hyperparameter tuning. Under random data-splitting approach, 30 different training and testing subsets are generated to assess the stability and robustness of model performance across different sample compositions. Random data partitioning is implemented at the level of constructed supervised learning instances, where each instance consisted of lagged input variables and their corresponding target values, forming distinct input–output pairs. The allocation of these instances to training and testing subsets is carried out entirely at random, without preserving sequential ordering or grouping temporally adjacent observations, while strictly maintaining original input–output correspondence within each constructed instance. The findings indicate that, for both simulated and real-world datasets, random data splitting resulted in improved predictive performance of both models, yielding lower error metrics compared to non-random splitting. These results suggested that random data splitting can enhance generalization and forecasting performance in time series applications with appropriate models. The study provides valuable empirical evidence on data partitioning strategies, supporting researchers and practitioners in making more informed decisions regarding model evaluation and selection in time series forecasting tasks.
Monkeypox Disease Classification Based on Skin Images Using Hierarchical Swin Transformer-Based Convolutional Neural Network Approach Ayu, Putu Desiana Wulaning; Sukket, Sasiwimol; Sutikno, Sutikno; Prihatini, Putu Manik; Pradipta, Gede Angga; Hostiadi, Dandy Pramana
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1313

Abstract

Monkeypox diagnosis can initially be conducted through expert physical examination based on characteristic lesions. However, laboratory confirmation using PCR is still essential, these tests are often hampered by limitations such as high costs, lengthy processing times, and a general lack of detailed symptom knowledge among patients. In light of these issues, image-based diagnostic methods offer a more efficient solution, given that monkeypox manifests as visible lesions on the skin that can be accurately detected using a deep learning. This study employs Transformer network-based deep learning for classifying skin diseases. To improve model robustness and mitigate the limitations of the relatively small dataset, we designed a comprehensive data augmentation pipeline that incorporates both positional and color transformations, including rotation, horizontal and vertical flipping, zooming, shearing, and brightness, contrast, hue, and saturation adjustments. Furthermore, a k-fold cross-validation strategy was employed, where the entire dataset was partitioned into k equal-sized folds to ensure a reliable and unbiased evaluation of the model performance. The Swin Transformer leverages advanced transformer network to analyze images, emphasizing hierarchical relationships within images. Swin Transformer enhances the convolutional Transformer architecture by substituting the standard multi-head-self-attention (MSA) mechanism with a shifted window-based MSA module It enhances efficiency over traditional transformer models by incorporating a shifted window mechanism, which reduces computational demands. The average global accuracy achieved was 0.99 (99%), which is further supported by the AUC values obtained for each disease category. The model achieved an AUC of 1.00 for chickenpox, cowpox, and hand-foot-mouth disease (HFMD), indicating excellent discriminative capability for these classes. Meanwhile, the remaining classes, including healthy skin, measles, and monkeypox, achieved AUC values of 0.99 and 0.98, respectively. These results demonstrate that the proposed Hierarchical Swin Transformer model provides highly reliable classification performance across all skin disease categories included in the dataset.
Iris Identification Using Resnet Iris Feature Extraction Architecture For Better Biometric Security Sama, Hendi; Tukino, Tukino; Siahaan, Mangapul; Titoni, Erica
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1166

Abstract

Iris recognition is widely acknowledged as one of the most reliable biometric modalities due to its high uniqueness, rich textural patterns, and long-term stability. Unlike other biometric traits, iris characteristics resist forgery, aging effects, and environmental variations, making it suitable for high-security applications. Recently, convolutional neural networks (CNNs) have been extensively applied in iris recognition to improve feature representation and classification accuracy. However, many CNN-based approaches still depend on conventional segmentation and handcrafted features, which reduce robustness under noisy data, illumination variations, occlusions, or unconstrained environments. To address these limitations, this study proposes an enhanced iris identification framework combining a modified T-Net for precise segmentation with deep residual feature extraction for improved discrimination. Unlike conventional systems focus mainly on classification, the proposed approach emphasizes segmentation-driven feature consistency, ensuring extracted features originate from accurately localized iris regions. This design enhances stability and reliability, particularly under challenging imaging conditions. The framework leverages transfer learning and efficient representation learning strategies, enabling high accuracy even with a limited labelled data. Evaluations on three benchmark datasets CASIA-IrisV4, IITD Iris Database, and UBIRIS.v2 covering both controlled and less-constrained acquisition scenarios. Results show that it achieves classification accuracy of up to 98.35%, while maintaining computational efficiency suitable for deployment. The proposed architecture offers a robust, data-efficient, and scalable solution for secure biometric authentication, with strong potential for real-world applications such as access control, identity verification, and high-security authentication systems.

Page 58 of 59 | Total Record : 583

 55   56   57   58   59 

Filter by Year

2020 2026


Reset
Filter By Issues
All Issue Vol 7, No 2: May 2026 Vol 7, No 1: January 2026 Vol 6, No 4: December 2025 Vol 6, No 3: September 2025 Vol 6, No 2: MAY 2025 Vol 6, No 1: JANUARY 2025 Vol 5, No 4: DECEMBER 2024 Vol 5, No 3: SEPTEMBER 2024 Vol 5, No 2: MAY 2024 Vol 5, No 1: JANUARY 2024 Vol 4, No 4: DECEMBER 2023 Vol 4, No 3: SEPTEMBER 2023 Vol 4, No 2: MAY 2023 Vol 4, No 1: JANUARY 2023 Vol 3, No 4: DECEMBER 2022 Vol 3, No 3: SEPTEMBER 2022 Vol 3, No 2: MAY 2022 Vol 3, No 1: JANUARY 2022 Vol 2, No 4: DECEMBER 2021 Vol 2, No 3: SEPTEMBER 2021 Vol 2, No 2: MAY 2021 Vol 2, No 1: JANUARY 2021 Vol 1, No 2: DECEMBER 2020 Vol 1, No 1: SEPTEMBER 2020 More Issue