Garuda - Garba Rujukan Digital

Clustering Key Performance Indicators using Convolutional Neural Networks Dimas Arditya Pinandhito; Herman Bedi Agtriadi; Luqman
Jurnal E-Komtek (Elektro-Komputer-Teknik) Vol 9 No 1 (2025)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v9i1.2322

Performance assessment based on Key Performance Indicators (KPI) is a crucial aspect in making strategic decisions in various industrial fields. Along with the development of artificial intelligence, the Convolutional Neural Network (CNN) method is starting to be applied to increase accuracy in KPI clustering. This research aims to analyze and compare the CNN approach in the KPI clustering process based on literature reviews from various scientific journals. The study results show that CNN is able to increase efficiency in KPI grouping with a better level of accuracy than conventional methods. This study is expected to provide deeper insight into the implementation of CNN in KPI analysis and open opportunities for further development in the future.

Evaluation of the Accuracy of the Naive Bayes Method in the Classification of Key Performance Indicators (KPIs) for Employees: Systematic Literature Review Chaerudin, Muhammad Farhan; Herman Bedi Agtriadi; Luqman
Jurnal E-Komtek (Elektro-Komputer-Teknik) Vol 9 No 1 (2025)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v9i1.2323

This study aims to evaluate the accuracy of Naive Bayes' method in classifying employee Key Performance Indicators (KPIs) through the Systematic Literature Review (SLR) approach. By collecting and analyzing reputable journals published between 2019 and 2024, this study examines the effectiveness of Naive Bayes in evaluating employee performance. The results of the study show that Naive Bayes is able to achieve a fairly high accuracy, which is between 84% to 90%, in classifying employee KPIs. However, this accuracy can vary depending on the complexity of the data used. Some research suggests that other methods such as Support Vector Machine (SVM) or Decision Tree may be superior in certain situations, especially when the data used is more complex or non-linear. In general, Naive Bayes remains a popular choice due to its ease of implementation and speed in delivering results. This study concludes that the selection of classification methods should be adjusted to the characteristics of the data and the purpose of the analysis to ensure optimal results.

Stasiun Pengisian Kendaraan Listrik Umum (SPKLU) Model Using GIS and Machine Learning Febryansyah Hans Arieyantho, Febry; Luqman; Agtriadi, Herman Bedi
ULTIMA InfoSys Vol 16 No 2 (2025): Ultima InfoSys : Jurnal Ilmu Sistem Informasi
Publisher : Universitas Multimedia Nusantara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31937/si.v16i2.4543

The adoption of electric vehicles in Indonesia is a strategic initiative supporting the national “Go Green” agenda and the Net Zero Emission target by 2060. As electric vehicle usage continues to grow, especially in West Java, well-distributed Stasiun Pengisian Kendaraan Listrik Umum (SPKLU) locations are increasingly important. Inefficient placement may lead to operational issues, reduced user convenience, and financial losses. This research develops such a model by integrating Geographic Information System (GIS) techniques with machine learning algorithms, specifically Multi-Layer Perceptron (MLP) and Support Vector Machine (SVM). Data preparation includes collecting spatial datasets such as road networks, housing, apartments, parking areas, public facilities from OSM, administrative boundaries from Geofabric, and existing SPKLU points from OpenData West Java. Proximity analysis is used to measure distances to key features, enabling classification of potential locations into Shared-Residential, Enroute, and Destination categories. These outputs are combined with socio-economic variables, including population density, income levels, vehicle ownership, household characteristics, education levels, and age distribution processed using Kernel Density Estimation (KDE). Results show that MLP performs best with 92.8% accuracy. The most influential variable is the productive-age population, minority population, unemployment, and total population. Overall, demographic factors play a dominant role in determining optimal SPKLU locations.

Tinjauan Literatur : Komparasi Certainty Factor dan Naïve Bayes untuk Diagnosis Penyakit Wijayanti , Anis; Luqman, Luqman; Agtriadi, Herman Bedi
Journal of Authentic Research Vol. 5 No. 2 (2026): May
Publisher : LITPAM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36312/b2f43293

Penelitian ini merupakan kajian literatur yang bertujuan menganalisis dan membandingkan metode Certainty Factor (CF) dan Naïve Bayes (NB) dalam pengembangan sistem diagnosis penyakit berbasis kecerdasan buatan. Kedua metode tersebut digunakan untuk meningkatkan akurasi dan efisiensi proses pengambilan keputusan dalam sistem pakar medis. Metode CF menerapkan pendekatan inferensi berbasis aturan untuk mengelola ketidakpastian melalui nilai tingkat keyakinan, sedangkan NB menggunakan prinsip probabilistik yang efektif dalam klasifikasi data medis berskala besar. Analisis dilakukan terhadap artikel ilmiah terpilih yang diterbitkan dalam lima tahun terakhir dan membahas penerapan kedua metode pada berbagai kasus diagnosis penyakit, termasuk diabetes, kanker, dan gangguan hormonal. Hasil kajian menunjukkan bahwa CF lebih sesuai untuk sistem pakar dengan basis pengetahuan eksplisit dan kebutuhan interpretasi tinggi, sedangkan NB menunjukkan performa yang lebih unggul pada data dengan distribusi statistik yang jelas dan volume besar. Perbandingan dilakukan berdasarkan tiga indikator utama, yaitu akurasi, kompleksitas implementasi, dan kemudahan interpretasi hasil. Temuan penelitian ini diharapkan menjadi acuan dalam pemilihan metode yang tepat untuk pengembangan sistem pakar diagnosis penyakit di bidang ilmu komputer dan kesehatan digital, sekaligus memperkuat penerapan kecerdasan buatan dalam mendukung diagnosis medis yang efisien dan berbasis data. This study presents a literature review aimed at analyzing and comparing the Certainty Factor (CF) and Naïve Bayes (NB) methods in the development of artificial intelligence–based disease diagnosis systems. Both methods are employed to enhance the accuracy and efficiency of decision-making processes in medical expert systems. The CF method applies a rule-based inference approach to handle uncertainty using confidence values, whereas NB employs a probabilistic approach that is effective for large-scale medical data classification. The analysis was conducted on selected scientific articles published within the last five years, discussing the implementation of both methods in various disease diagnosis cases, including diabetes, cancer, and hormonal disorders. The results indicate that CF is more suitable for expert systems requiring explicit knowledge representation and high interpretability, while NB demonstrates superior performance in scenarios involving large data volumes with well-defined statistical distributions. The comparison focuses on three main indicators: accuracy, implementation complexity, and result interpretability. The findings are expected to serve as a reference for selecting appropriate methods in developing medical expert diagnosis systems within the fields of computer science and digital health, and to strengthen the application of artificial intelligence in supporting data-driven and efficient medical diagnostics.

Klasifikasi Citra Magnetic Resonance Imaging (MRI) Penyakit Tumor Otak Menggunakan Metode Convolutional Neural Network (CNN) Dengan Arsitektur VGG-16 Dan Xception Melania Erika Andolo; Herman Bedi Agtriadi; Meilia Nur Indah Susanti
Jurnal E-Komtek Vol 10 No 1 (2026)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v10i1.2974

rain tumors are critical medical conditions that require early diagnosis to improve treatmentoutcomes. Magnetic Resonance Imaging (MRI) is widely utilized for brain tumor detection due to itsability to produce high-resolution images of soft tissues. Nevertheless, manual interpretation of MRIimages presents several challenges, including time inefficiency and variability among observers. Toaddress these issues, this study applies the Convolutional Neural Network (CNN) approach usingVGG-16 and Xception architectures to classify brain tumor MRI images and to evaluate theirperformance comparatively. The dataset comprises 2,877 MRI images categorized into four classes:glioma tumor, meningioma tumor, pituitary tumor, and no tumor. Preprocessing stages includeresizing images to 224×224 pixels and dividing the dataset into training, validation, and testing setswith a ratio of 80:10:10. Model performance is assessed using accuracy, precision, recall, and F1-scoremetrics. Experimental results indicate that the VGG-16 architecture achieves an accuracy of 92%, whilethe Xception architecture records an accuracy of 91%.

Comparative Analysis of PCC Load Balancing and Quality of Service (QoS) Queue Tree Methods on Mikrotik for Bandwidth Optimization Duliades Mirsal; Herman Bedi Agtriadi
Jurnal E-Komtek Vol 10 No 1 (2026)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v10i1.3170

The increasing demand for fast, stable, and efficient Internet connectivity requires optimal bandwidth management, especially in PLN's operational networks. This study analyzes and compares two MikroTik bandwidth management methods, namely Per Connection Classifier (PCC) Load Balancing and Quality of Service (QoS) Queue Tree, to optimize network performance. An experimental approach was conducted under three traffic conditions: low, medium, and high. The evaluation focused on key network parameters, including throughput, latency, packet loss, and connection stability. The results indicate that PCC Load Balancing achieves higher throughput and more effective traffic distribution across multiple Internet Service Providers (ISPs), making it suitable for networks with many general users. In contrast, QoS Queue Tree provides lower latency and jitter while ensuring better traffic prioritization for critical services such as SCADA systems and internal communications. Therefore, PCC is more effective in maximizing bandwidth utilization, whereas QoS Queue Tree offers superior service quality and network stability. A hybrid implementation of both methods is recommended to achieve optimal efficiency and performance in PLN operational environments.

CT-Scan Image Segmentation of Liver Cancer Using the Active Contour Method Nurazizah Nurazizah; Dwina Kuswardani; Herman Bedi Agtriadi
Jurnal E-Komtek Vol 10 No 1 (2026)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v10i1.3273

Liver cancer is one of the most common types of cancer both in Indonesia and throughout the world. The increase in liver cancer cases is thought to be related to the increase in hepatitis B and C virus infections. According to statistical data from Globocan, in 2022 there will be 866,000 cases of liver cancer recorded worldwide, with 758,000 deaths due to this disease. To overcome this challenge, image processing plays an important role in improving image quality and performing segmentation to separate objects based on certain characteristics. In this research, the Contrast Limited Adaptive Histogram Equalization (CLAHE) and Contrast Stretching methods are used to improve image quality, while the Active Contour method is applied for image segmentation. Validation of segmentation results was carried out using Receiver Operating Characteristic (ROC) calculations. Testing was conducted on 21 CT-scan images of liver cancer, yielding an accuracy of 97.16%, a sensitivity of 75.69%, and a specificity of 98.85%. This research was conducted using the MATLAB application (R2015a). These findings demonstrate the effectiveness of the methods used in supporting the diagnosis and treatment of liver cancer

Implementation of The K-Nearest Neighbors (KNN) Algorithm in The Process of Student Graduation Prediction (Case Study of The Bachelor of Informatics Engineering Program, PLN Institute of Technology, Jakarta) Andi Abd. Jalil. L; Herman Bedi Agtriadi; Meilia Nur Indah; Rakhmadi Ifansyah Putra
Jurnal E-Komtek Vol 10 No 1 (2026)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v10i1.3299

Student graduation is one of the key indicators in a university’s Internal Quality Assurance System (SPMI). Based on data from the Bachelor of Informatics Engineering program, out of 305 students from the 2016 cohort, 227 graduated on time and 78 graduated late. This study aims to predict student graduation using the K-Nearest Neighbors (KNN) algorithm. The research stages include data collection and division for training and testing, parameter determination with K=3, and distance calculation between data points. The results show that the KNN model with parameter K=3 achieved an accuracy rate of 90% in predicting student graduation. This demonstrates that the KNN method is effective in predicting student graduation outcomes.

Tinjauan Literatur : Komparasi Certainty Factor dan Naïve Bayes untuk Diagnosis Penyakit Wijayanti , Anis; Luqman, Luqman; Agtriadi, Herman Bedi
Journal of Authentic Research Vol. 5 No. 2 (2026): May
Publisher : LITPAM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36312/b2f43293

Penelitian ini merupakan kajian literatur yang bertujuan menganalisis dan membandingkan metode Certainty Factor (CF) dan Naïve Bayes (NB) dalam pengembangan sistem diagnosis penyakit berbasis kecerdasan buatan. Kedua metode tersebut digunakan untuk meningkatkan akurasi dan efisiensi proses pengambilan keputusan dalam sistem pakar medis. Metode CF menerapkan pendekatan inferensi berbasis aturan untuk mengelola ketidakpastian melalui nilai tingkat keyakinan, sedangkan NB menggunakan prinsip probabilistik yang efektif dalam klasifikasi data medis berskala besar. Analisis dilakukan terhadap artikel ilmiah terpilih yang diterbitkan dalam lima tahun terakhir dan membahas penerapan kedua metode pada berbagai kasus diagnosis penyakit, termasuk diabetes, kanker, dan gangguan hormonal. Hasil kajian menunjukkan bahwa CF lebih sesuai untuk sistem pakar dengan basis pengetahuan eksplisit dan kebutuhan interpretasi tinggi, sedangkan NB menunjukkan performa yang lebih unggul pada data dengan distribusi statistik yang jelas dan volume besar. Perbandingan dilakukan berdasarkan tiga indikator utama, yaitu akurasi, kompleksitas implementasi, dan kemudahan interpretasi hasil. Temuan penelitian ini diharapkan menjadi acuan dalam pemilihan metode yang tepat untuk pengembangan sistem pakar diagnosis penyakit di bidang ilmu komputer dan kesehatan digital, sekaligus memperkuat penerapan kecerdasan buatan dalam mendukung diagnosis medis yang efisien dan berbasis data. This study presents a literature review aimed at analyzing and comparing the Certainty Factor (CF) and Naïve Bayes (NB) methods in the development of artificial intelligence–based disease diagnosis systems. Both methods are employed to enhance the accuracy and efficiency of decision-making processes in medical expert systems. The CF method applies a rule-based inference approach to handle uncertainty using confidence values, whereas NB employs a probabilistic approach that is effective for large-scale medical data classification. The analysis was conducted on selected scientific articles published within the last five years, discussing the implementation of both methods in various disease diagnosis cases, including diabetes, cancer, and hormonal disorders. The results indicate that CF is more suitable for expert systems requiring explicit knowledge representation and high interpretability, while NB demonstrates superior performance in scenarios involving large data volumes with well-defined statistical distributions. The comparison focuses on three main indicators: accuracy, implementation complexity, and result interpretability. The findings are expected to serve as a reference for selecting appropriate methods in developing medical expert diagnosis systems within the fields of computer science and digital health, and to strengthen the application of artificial intelligence in supporting data-driven and efficient medical diagnostics.

Clustering for Motivating New Student Admissions in Study Program Selection: Systematic Literature Review Afif Dwi Kurniawan; Luqman; Herman Bedi Agtriadi
Jurnal E-Komtek Vol 9 No 2 (2025)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/e-komtek.v9i2.2510

This research aims to evaluate clustering in new student admissions in determining effective strategies, to help prospective students in choosing study programs that match the interests and potential of prospective new students. Clustering as a machine learning technique to group data that has similarities, is increasingly used in the field of education to support the decision-making process. This Systematic Literature Review (SLR) examines the application of clustering methods in new student admissions, especially in recommending the right study program. By analyzing 10 studies in applying clustering methods that are often used, to determine the main factors that influence the selection of courses, as well as their impact on student satisfaction in choosing courses and optimal academic results. The results of this study provide insight into strategies for the admissions team in optimizing marketing, so that there is a more effective alignment between student profiles and study program characteristics.

Title

Found 21 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 21 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 21 Documents
Search