Article Per Year (5 Year)
p-Index From 2021 - 2026
1.261
P-Index
This Author published in this journals
All Journal TELKOMNIKA (Telecommunication Computing Electronics and Control) Telematika Jurnal Tekno Insentif Indonesian Journal of Applied Informatics Journal of Applied Computer Science and Technology (JACOST)
Roni Habibi
Universitas Logistik dan Bisnis Internasional
Computer Science & IT Education

Published : 6 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 6 Documents
Search

 1 
Pendekatan Supervised Learning untuk Diagnosa Kehamilan Fahira Fahira; Zian Asti Dwiyanti; Roni Habibi
Jurnal Tekno Insentif Vol 17 No 2 (2023): Jurnal Tekno Insentif
Publisher : Lembaga Layanan Pendidikan Tinggi Wilayah IV

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36787/jti.v17i2.1102

Abstract

Abstrak Dalam studi ini dilakukan evaluasi performa dari beberapa algoritma machine learning dalam mendiagnosis kehamilan. Tujuan dari studi ini adalah untuk menemukan algoritma yang paling efektif antara decision tree dengan random forest dalam mendiagnosis kehamilan. Studi ini juga bertujuan untuk memberikan pengetahuan yang lebih baik mengenai bagaimana teknik pembelajaran mesin dapat digunakan dalam proses diagnosa kehamilan dan memberikan informasi yang berguna bagi para dokter dan peneliti dalam membuat keputusan yang tepat dalam mendiagnosis kehamilan. Dari hasil evaluasi dapat disimpulkan bahwa model Random Forest dengan menggunakan dataset yang seimbang dan metode Gini memiliki akurasi terbaik sebesar 81%. Hal ini menunjukkan bahwa menggunakan dataset yang seimbang dapat meningkatkan performa dalam mendiagnosis kehamilan. Algoritma Random Forest merupakan metode yang sering digunakan dalam proses pengklasifikasian karena kinerjanya yang baik. Algoritma ini bekerja dengan membuat pohon keputusan yang digunakan untuk membuat prediksi. Pada studi ini, kami menggunakan algoritma Random Forest dengan metode Gini untuk melakukan prediksi kehamilan. Abstract This study evaluates the performance of several machine learning algorithms in diagnosing pregnancy. The purpose of this study is to find the most effective algorithm between decision trees and random forests in diagnosing pregnancy. This study also aims to provide better knowledge about how machine learning techniques can be used in the process of diagnosing pregnancy and provide useful information for doctors and researchers in making the right decisions in diagnosing pregnancy. From the evaluation results it can be concluded that the Random Forest model using a balanced dataset and the Gini method has the best accuracy of 81%. This shows that using a balanced dataset can improve performance in diagnosing pregnancy. Random Forest Algorithm is a method that is often used in the classification process because of its good performance. This algorithm works by creating a decision tree that is used to make predictions. In this study, we used the Random Forest algorithm with the Gini method to predict pregnancy.
SLR Systematic Literature Review: Metode Penilaian Kinerja Karyawan Menggunakan Human Performance Technology Roni Habibi; Artha Glory Romey Manurung
Journal of Applied Computer Science and Technology Vol 4 No 2 (2023): Desember 2023
Publisher : Indonesian Society of Applied Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52158/jacost.v4i2.511

Abstract

Performance appraisal is an activity conducted by groups or individuals in a company with the aim of evaluating employee performance based on specific indicators. The purpose of this evaluation is to determine the level of employee performance, provide fair job opportunities, and enhance employee motivation and performance. The objective of this research is to review the literature that employs the Systematic Literature Review (SLR) method, while considering several aspects such as previous research on employee performance appraisal and human performance technology. Initially, the search was limited to 450 journals. Then, a review and further screening were conducted, resulting in 161 remaining journals using inclusion and exclusion criteria, yielding 17 journals that met the requirements. This research was conducted over several years to assess employee performance. Therefore, this systematic literature review is carried out to identify the methods used for employee performance appraisal and mentions that optimization algorithms and fuzzy comprehensive methods are used as evaluation methods to assess performance appraisal methods.
MDI and PI XGBoost regression-based methods: regional best pricing prediction for logistics services Agus Purnomo; Aji Gautama Putrada; Roni Habibi; Syafrianita Syafrianita
TELKOMNIKA (Telecommunication Computing Electronics and Control) Vol 22, No 5: October 2024
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/telkomnika.v22i5.26037

Abstract

The logistics industry in Indonesia, with PT Pos Indonesia as the dominant player, is confronted with intense price competition. The challenge lies in establishing the most favorable price for regional logistics services in every region, with the aim of gaining a competitive edge and augmenting revenue. This intricate task encompasses local market conditions, competition, customer preferences, operational costs, and economic factors. To address this complexity, this study proposes the utilization of machine learning for price prediction. The price prediction model devised incorporates the extreme gradient boosting regression (XGBR), support vector machine (SVM), random forest, and logistics regression algorithms. This research contributes to the field by employing mean decrease in impurity (MDI) and permutation importance (PI) to elucidate how machine learning models facilitate optimal price predictions. The findings of this study can assist company management in enhancing their comprehension of how to make informed pricing decisions. The test results demonstrate values of 0.001, 0.005, 0.458, 0.009, and 0.9998. By employing machine learning techniques and explanatory models, PT Pos Indonesia can more accurately determine optimal prices in each region, bolster profits, and effectively compete in the expanding regional market.
Systematic Review of Supervised Learning Models for Network Flood Detection (NFD): Trends, Performance Evaluation, and Implementation Insights Roni Habibi; Naufal Dekha Widana
Telematika Vol 18, No 2: August (2025)
Publisher : Universitas Amikom Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35671/telematika.v18i2.3183

Abstract

Due to the growing volume, speed, and sophistication of malicious traffic, Network Flood Detection (NFD), especially in the context of Distributed Denial of Service (DDoS) assaults, continues to be a crucial challenge in contemporary network security.  Supervised machine learning has been widely used to enhance the precision, scalability, and real-time detection capabilities of NFD systems.  However, current research reveals inconsistent results on the optimal supervised learning algorithm, mostly because of differences in datasets, feature engineering methods, assessment criteria, and deployment settings.  In order to assess supervised learning models applied to NFD, this study intends to do a Systematic Literature Review (SLR) utilizing the PRISMA framework. A structured search was performed via Scopus, IEEE Xplore, SpringerLink, and ScienceDirect, encompassing papers from 2019 to 2025.  40 primary papers and 16 additional articles were found to be appropriate for synthesis after an initial dataset of 516 research was reviewed using predetermined inclusion and exclusion criteria.  Algorithms, datasets, evaluation criteria, feature selection techniques, and deployment characteristics were all incorporated in the data extraction process. According to the review, models like Random Forest, XGBoost, K-Nearest Neighbor, and Support Vector Machine regularly perform well, with accuracy ranging from 92% to 99%, depending on preprocessing methods and dataset features.  Common problems highlighted include dataset imbalance, lack of real-time adaptation, and insufficient generalization to unforeseen assault types. The results show that supervised learning is still a promising method for NFD, particularly when combined with balanced datasets, hybrid or ensemble model techniques, and optimized feature engineering.  To increase real-time resilience against changing network threats, further research is urged to incorporate deep learning, lightweight edge models, and adaptive learning frameworks.
Pengaruh Metode Seleksi Fitur terhadap Akurasi Model SVM dalam Klasifikasi Customer Churn pada Perusahaan Telekomunikasi Mayke Andani Rohmaniar; Roni Habibi; Syafrial Fachri Pane
IJAI (Indonesian Journal of Applied Informatics) Vol 9, No 1 (2024)
Publisher : Universitas Sebelas Maret

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20961/ijai.v9i1.92983

Abstract

Abstrak:Penelitian ini menganalisis pengaruh metode seleksi fitur terhadap akurasi model Support Vector Machine dalam memprediksi pelanggan di industri telekomunikasi. Empat metode seleksi fitur (Correlation Matrix, PCA, dan GA) dan empat kernel (Linear, Polynomial, RBF, dan Sigmoid) dibandingkan menggunakan dataset pelanggan telekomunikasi dari Kaggle dengan 7043 entri dan 33 fitur. Metodologi CRISP-DM digunakan, meliputi Pemahaman Bisnis, Pemahaman Data, Persiapan Data, Pemodelan, Evaluasi, dan Implementasi. Hasil penelitian menunjukkan bahwa metode seleksi fitur menggunakan Correlation Matrix dengan kernel Linear memberikan kinerja terbaik. Model ini mencapai akurasi tertinggi sebesar 92,48%, dengan precision 0,93, recall 0,97, dan f1-score 0,95. Metode seleksi fitur lainnya, seperti PCA dan GA, memberikan hasil yang lebih rendah dibandingkan dengan Correlation Matrix. Implementasi model prediksi yang akurat diharapkan dapat membantu perusahaan telekomunikasi mengembangkan strategi retensi pelanggan yang lebih efektif.=================================================Abstract:This study examines the impact of various feature selection methods on the accuracy of the Support Vector Machine (SVM) model in predicting customer behavior within the telecommunications sector. Specifically, the research compares four feature selection techniques: Correlation Matrix, Principal Component Analysis (PCA), and Genetic Algorithm (GA). Additionally, it evaluates the performance of four SVM kernels: Linear, Polynomial, Radial Basis Function (RBF), and Sigmoid. Utilizing a telecom customer dataset from Kaggle, which comprises 7043 entries and 33 features, the study adheres to the CRISP-DM methodology. This methodology includes phases such as Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Implementation. The findings indicate that the Correlation Matrix feature selection method, when paired with the Linear kernel, provides the best performance. This particular configuration achieves the highest accuracy rate of 92.48%, along with a precision score of 0.93, a recall score of 0.97, and an F1-score of 0.95. In contrast, other feature selection methods, such as PCA and GA, result in lower performance metrics. These findings underscore the effectiveness of the Correlation Matrix and Linear kernel combination in enhancing the predictive accuracy of SVM models.
Systematic Literature Review: Tren dan Tantangan Machine Learning pada Sistem Rekomendasi Kursus Daring Roni Habibi; Ryaas Ishlah Ramadhan
Journal of Applied Computer Science and Technology Vol. 6 No. 2 (2025): Desember 2025
Publisher : Indonesian Society of Applied Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52158/ct1kr873

Abstract

This study presents a Systematic Literature Review (SLR) on the application of machine learning in online course recommendation systems. The aim is to map research trends, methodological approaches, and challenges in developing AI-based recommendation systems for online education. A total of 40 Scopus-indexed articles published between 2020 and 2025 were analyzed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines, using the Watase UAKE tool for literature selection. The findings reveal that deep learning and hybrid models are the most dominant approaches, with a significant increase observed during 2022–2024. China contributes 57.5% of the studies, followed by India and Taiwan, indicating a strong research concentration in Asia. Combined architectures such as CNN–LSTM–ResNet achieved the highest accuracy (99.2%), while Reinforcement Learning (RL) and Deep Reinforcement Learning (DRL) are emerging as adaptive approaches for course recommendation. The main challenges identified include real-time adaptability, computational efficiency, and model transparency. The main contribution of this paper is to provide a comprehensive map of current research and outline future directions toward adaptive, efficient, and explainable online course recommendation systems.T
Co-Authors Agus Purnomo Aji Gautama Putrada Artha Glory Romey Manurung Fahira Fahira Mayke Andani Rohmaniar Naufal Dekha Widana Ryaas Ishlah Ramadhan Syafrial Fachri Pane Syafrianita Syafrianita Zian Asti Dwiyanti