cover
Contact Name
Christian Harito
Contact Email
christian.harito@binus.edu
Phone
+6221-5350660
Journal Mail Official
aagung@binus.edu
Editorial Address
Universitas Bina Nusantara Jl. Kebon Jeruk Raya No.27 Kebon Jeruk, Jakarta Barat 11530
Location
Kota adm. jakarta barat,
Dki jakarta
INDONESIA
Engineering, Mathematics and Computer Science Journal (EMACS)
ISSN : -     EISSN : 26862573     DOI : https://doi.org/10.21512/emacs
Engineering, MAthematics and Computer Science (EMACS) Journal invites academicians and professionals to write their ideas, concepts, new theories, or science development in the field of Information Systems, Architecture, Civil Engineering, Computer Engineering, Industrial Engineering, Food Technology, Computer Science, Mathematics, and Statistics through this scientific journal.
Articles 199 Documents
Effect of Price Volatility on LSTM Lookback Windows in Indonesian Banking Stocks Joan Christina Bahagiono
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 1 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i1.15388

Abstract

This study aims to explore how stock price volatility influences the sequence length or also known as lookback window hyperparameter of LSTM. This study uses a comparative approach to determine the relationship between stock prices volatility and the best lookback window to achieve the lowest error rate of an LSTM model in predicting stock prices. Nine selected stocks in the banking sector of Indonesia Stock Exchange were compared, ranging from relatively stable to volatile. The banking sector was used as it contains multiple stocks under the same sector that varies in price movement volatility. An aggregation was also conducted to produce grouped results. The results of this study highlighted the importance of hyperparameter tuning in LSTM especially in the lookback window hyperparameter. Shorter LSTM lookback window is well suited in low volatility stocks, with the lowest mean squared error rate of 0.030782 observed in this study at the 42 trading days lookback period. In contrast to that, highly volatile stocks exhibit a different pattern, where longer lookback period improves LSTM prediction performance, as demonstrated in this study through a 0.016001 mean squared error at the 252 trading days lookback period. The findings imply that high volatility stocks require longer temporal memory in the LSTM to capture complex and irregular price movements, whereas low volatility stocks are better modelled using shorter and more recent information.
Design and Implementation of a Web Service-Based Middleware Application for PDDIKTI Reporting at XYZ University Harkat Christian Zamasi
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 1 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i1.15355

Abstract

Reporting of academic data to the Higher Education Database (PDDIKTI) is an obligation for every university in Indonesia that must be done accurately and timely in accordance with government regulations. However, in practice, the reporting process is still largely manual via the Neo Feeder application. This causes problems in efficiency, data consistency and possible delays in reporting, particularly in universities with large volumes of academic data. Although PDDIKTI has provided web services facilities, their utilization as part of the Extract, Transform, and Load (ETL) middleware architecture has not been optimally implemented. However, the integration of internal academic information system with PDDIKTI still has problems related to the difference of data structure, business rules and technology platform used and the absence of a generally replicable middleware model. This study aims to design and implement the ETL middleware architecture model based on web services Neo Feeder as a solution for the automation of PDDIKTI data reporting. The research method used is applied research (applied research) with an approach to information systems engineering, which consists of the stages of needs analysis, design of system architecture, implementation and evaluation of prototypes in a case study of XYZ University. The results of this study indicate that the proposed ETL middleware architecture model can integrate internal academic information systems with PDDIKTI in an automatic, structured and controlled manner, thus increasing the efficiency and timeliness of academic data reporting. The resulting architecture model and prototype are generic and can be replicated by other universities as a reference for implementing integrated PDDIKTI data reporting in line with the One Data Higher Education policy.
Leveraging Artificial Intelligence (AI) Methods for Non-Small Cell Lung Cancer (NSCLC) Detection: A Review Faisal Asadi; Ashraf Alif Adillah; Jonathan Lucas Fontana; Michael Dimas Chrispradipta; Mousa Khalil Mousa Ayesh; Wairanatha Halim
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 1 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i1.15454

Abstract

One of the main causes of cancer-death that significant to global public health concern is lung cancer. NSCLC has been categorized as a burden disease, with an estimated reaching 85% of all lung cancer cases around the world. The problem was that NSCLC disease could only be detected when the disease has grown at a late stage. Therefore, AI technology is also being implemented to handle NSCLC disease. This review discusses how AI has played a role in treating NSCLC disease in the last five years of research journals that collected 5 years between 2019-2024. The database resources are from PubMed, Scopus, and Google Scholar. The process of selecting journal papers was analyzed based on an in-depth understanding of NSCLC disease journals as considered an inclusion criterion. This review used the PRISMA to analysis and review 17 journals. After carrying out the analysis process on the AI-NSCLC journals, we found that AI has been able to help humans respond to cases of NSCLC patients, starting from the detection stage, comprehensive diagnosis, and providing treatment recommendations. Treatments of NSCLC tend to be more personalized and could run more effectively and efficiently based on medical images input into the AI model. However, considering the urgency and vulnerability of the application of these AI models, which will be directly related to human health, the medical images dataset is also quite limited and the biggest challenge for AI-NSCLC.
The Paradox of Web Service Composition and Load Balancing: Theoretical Compatibility vs. Business Reality Maksymilian Iwanow
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 1 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i1.15594

Abstract

Web services have become a fundamental building block of modern IT infrastructures. They underpin both: internal system integration and the delivery of complex business functionalities – and as organizations continue to shift toward service-oriented or microservice-based architectures, managing service interactions well is no longer an option. Composition and load balancing have grown into critical concerns for any company serious about keeping production systems running reliably at scale. This paper explores where service composition and load balancing conceptually overlap – and, perhaps more interestingly, where they contradict each other. The tension becomes visible when these mechanisms are treated as multi-criteria optimization problems rather than as pragmatic, cost-driven business decisions. To examine this, the study combines a structured literature review with informal participant observation carried out in a real enterprise environment. What emerges from the analysis is that service composition – broadly, the integration of multiple sub-services to fulfill a specific functional goal – frequently runs into load balancing, which distributes incoming requests across service instances to make efficient use of available resources. In practice, though, this intersection tends to be poorly understood. Academic literature and industry practice alike treat both mechanisms inconsistently, often without acknowledging the tensions between them. This paper tries to cut through that ambiguity by combining a close reading of the literature with hands-on professional perspective. The result is a synthesis that places itself somewhere between theory and practice. In spite of definitive answers, the paper is also intended as a starting point for a broader conversation – one that feels overdue, given how many relevant questions remain underexplored.
Probabilistic Liquefaction Potential Mapping in South Lebong Based on PSHA-Derived PGA Diefi Deayuzeta; Lindung Zalbuin Mase; Fepy Supriani; Tri Bintang Pratitis; Rena Misliniyati; Khairul Amri
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 1 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i1.15761

Abstract

This study evaluates liquefaction potential in South Lebong Subdistrict, Lebong Regency, Bengkulu Province, which is located in an active tectonic region influenced by the Sumatra Fault system and dominated by alluvial deposits. A probabilistic approach using the Probabilistic Seismic Hazard Analysis method to determine Peak Ground Acceleration (PGA) values for 10% and 2% probabilities of exceedance in 50 years, representing moderate to extreme earthquake conditions. South Lebong District is dominated by PGA values greater than 0.8g, particularly under the 2% probability scenario, indicating relatively high seismic hazard in the study area. Under the 10% probability scenario, several locations still show PGA values ranging from 0.4g to 0.8 g. The obtained PGA values were subsequently used to evaluate liquefaction potential through calculations of the Cyclic Stress Ratio (CSR), Cyclic Resistance Ratio (CRR), and Factor of Safety (FS). Additionally, susceptibility was assessed using the Liquefaction Potential Index (LPI) method. The results indicate that higher PGA values under the 2% probability scenario increase seismic loading intensity and liquefaction susceptibility within the study area. Based on the LPI classification, liquefaction potential under the 10% probability scenario is generally categorized as low to moderate, whereas under the 2% probability scenario, several locations shift into the moderate to severe liquefaction category. These findings indicate that liquefaction susceptibility in the South Lebong District is strongly influenced by the interaction between earthquake loading intensity and local geotechnical conditions. Therefore, the results of this study can support earthquake hazard mitigation planning and the development of safer areas that are less vulnerable to liquefaction hazards.
Integrating Finite Element Analysis and Machine Learning to Predict the Bearing Capacity of Strip Footings on Slopes Kurniawan, Aditya Dwi; Mase, Lindung Zalbuin; Fikri, Muharram Nur; Misliniyati, Rena; Fitriansyah, Aidil
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 2 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i2.16161

Abstract

Predicting the ultimate bearing capacity (qult) of strip footings on slopes remains a major challenge in geotechnical engineering, as classical methods were developed for level ground and lose reliability for complex slope-foundation geometries. Although finite element analysis (FEA) provides better accuracy, its high computational cost limits large-scale parametric studies. This study proposes a hybrid FEA–machine learning (ML) framework to estimate qult of strip footings on slopes, overcoming the accuracy limitations of existing analytical solutions for different slope-foundation configurations. The dataset of 600 finite element simulations was developed under the Mohr-Coulomb plane strain constitutive framework. Six variables were examined: unit weight (γ), cohesion (c), friction angle (φ), applied load (P), foundation width (B), and embedment depth (Df). Seven predictive models were developed: multiple linear regression, polynomial regression, support vector regression, decision trees, random forests, k-nearest neighbors, and extreme gradient boosting (XGBoost). Model performance was assessed using R², RMSE, MAPE, and the a20 index, with R² and RMSE as the primary ranking criteria, while Shapley Additive Explanations (SHAP) were applied to interpret feature contributions. XGBoost has the highest prediction accuracy on both the training and test datasets. It is followed by Support Vector Regression (SVR). The most influencing parameter in all seven models was the foundation depth (Df), followed by the friction angle (φ) and the foundation width (B), while the slope angle consistently decreased the predicted bearing capacity. The results confirm the accuracy, interpretability, and computational efficiency of the integrated FEA-ML approach as an alternative to traditional bearing capacity analysis.
Enhancing K-Nearest Neighbor Algorithm on Network Traffic Classification using Recursive Feature Elimination Amalia, Andika Elok; Bahari, Alwin; Bayuaji, Luhur
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 2 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i2.14277

Abstract

Network traffic classification is a category in network traffic monitoring analysis that has a very important function in network management analysis. By identifying the type of traffic flowing through a network, administrators can allocate bandwidth more effectively, maintain service quality, and detect abnormal activity, which makes it a fundamental part of network monitoring. One of the ways to do it is by machine learning both unsupervised and supervised. This study implemented supervised machine learning, namely K-Nearest Neighbor to network traffic dataset with two different schemas, with and without feature selection. These two schemas are compared directly so that any difference in results can be attributed to the feature selection process. It aims to clarify the impact of Recursive Feature Elimination to network traffic classification using K-Nearest Neighbor Algorithm. It classifies data with features to 17 target classes. Previous studies state that the use of feature selection in some techniques can improve its accuracy. Recursive Feature Elimination is chosen and result that the accuracy is increased from 0.66 to 0.76. Another thing that this study shows is that there is no pattern where the more data points from a class will result in better classification performance in that class. In other words, the number of samples available for a class does not guarantee that the class will be classified more accurately. This research clarifies that feature selection can improve classification performance. Selecting the most relevant features also reduces the amount of irrelevant information the algorithm must process, which makes the classification more efficient and reliable,
Analysis of K-Means Clustering Implementation for New Student Segmentation Based on School of Origin and Study Program Pradini, Risqy Siwi; Yulianti, Yulianti; Anshori, Mochammad
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 2 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i2.14486

Abstract

Increasing Competition among higher education institutions requires a deeper understanding of prospective students characteristics to develop more effective promotional strategies and student admission policies. The K-Means Clustering Algorithm is used in this study to divide recently admitted students into discrete groups according to their chosen course of study, kind of previous school, and place of origin. The Davies-Bouldin Index (DBI) and the Silhouette Score were used to assess the clustering performance's efficacy, and the Elbow Method was used to determine the ideal number of clusters. The Experimental result indicated that the optimal segmentation consisted of two clusters. The first cluster was primarily composed of students graduating from general senior high schools (SMA), who tended to enroll in highly demanded health-related study programs. Conversely, the second cluster exhibited a more heterogeneous distribution concerning study program preferences and was predominantly constituted of students hailing from Islamic senior high schools (MA) and vocational high schools (SMK). The cluster analysis has been found to give a Silhouette Score of 0.1774 and a Davies–Bouldin Index of 2.1806, meaning that although the separation between clusters is still inadequate, it is good enough to reveal some distinct patterns in the traits of the new students admitted. The results of this research can be used as a useful benchmark for developing marketing and partnership policies with secondary schools and making data-based decisions-making in the new student admission process.
Deriving Closed-Form Expressions for Multi-Level Line Series within the Gradual Summation Method Framework via Regularity Theory Goenawan, Stephanus Ivan
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 2 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i2.15322

Abstract

The present research develops closed-form mathematical models for multi-level line series using the fundamental concepts of the Theory of Regularity. This research focuses on developing general analytical formulas for first-level series in the Gradual Summation Method (GSM) Table and subsequently extending them to higher levels of the series. Using regular pattern analysis derived from the fundamental principles of the Theory of Regularity, we create general notation and precise formulas. Our findings indicate that our proposed multi-level line series model can accurately compute both ordered sequences from the GSM Table and any other real number series using closed-form analytical equations rather than relying on traditional recursive methods. The primary benefit of our research is that we provide a mathematically rigorous model that replaces repeated summations with analytical formulas. Therefore, our model provides a strong theoretical and practical method for computing hierarchical data structures through software implementation. Our model serves as an analytical substitute for traditional loop-based computational algorithms. Additionally, we describe the structural properties, boundary conditions, and operator functions that define multi-level evaluations. Our findings help connect the field of discrete sequence analysis with practical algorithm design. Future applications may utilize the precise formulas we developed to improve multi-level aggregation tasks across many different types of computing environments.
Optimizing Deep Learning using Feature Selection for Accurate Health Insurance Claim Classification Wardhani, Anindya Khrisna; Irwan Syah; Sudra, Rano Indradi; Lakhmudien, Lakhmudien; Bireedan, Abdulmunem Abdullah Mohemmed
Engineering, MAthematics and Computer Science Journal (EMACS) Vol. 8 No. 2 (2026): EMACS
Publisher : Bina Nusantara University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21512/emacsjournal.v8i2.15796

Abstract

Accurate prediction of health insurance claims is essential for improving hospital service quality, financial management, and operational decision-making. Health insurance claim data contains multiple attributes that may affect the performance and efficiency of classification models. Data mining techniques, particularly deep learning, are widely applied because of their ability to identify complex patterns in large datasets. However, the use of numerous attributes can increase model complexity and computational requirements. Therefore, feature selection can be applied to reduce data dimensionality while retaining relevant information for the classification process. This study aims to optimize a deep learning model using the Weight by User Specification method as a feature selection technique for classifying BPJS Health insurance claims. The dataset consists of 2,388 BPJS Health insurance claim records from Kumala Siwi Hospital, with variables including total rate, hospital rate, profit, diagnosis, and INACBG. The research process follows data preprocessing, data transformation, deep learning classification, feature selection, and model evaluation using a confusion matrix. Two classification models were evaluated: deep learning without feature selection and deep learning combined with Weight by User Specification. The results show that the deep learning model without feature selection achieved an accuracy of 98.87%, while the optimized model achieved 99.20%, representing an improvement of 0.33%. The classification error also decreased from 1.13% to 0.80%. These findings indicate that integrating feature selection with deep learning can improve classification accuracy and reduce classification error in BPJS Health insurance claim prediction. Therefore, weight by user specification combined with deep learning provides an effective approach for health insurance claim classification.