Claim Missing Document
Check
Articles

Model Hybrid Random Forest dan Information Gain untuk Meningkatkan Performa Algoritma Machine Learning pada Deteksi Malicious Software Rafrastara, Fauzi Adi; Ghozi, Wildanil; Sani, Ramadhan Rakhmat; Handoko, L. Budi
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol 6, No 2 (2024): September
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36499/jinrpl.v6i2.11216

Abstract

Evolusi malware atau perangkat lunak berbahaya semakin meningkatkan kekhawatiran, menyerang tidak hanya komputer tetapi juga perangkat lain seperti smartphone. Malware kini tidak hanya berbentuk monomorfik, tetapi telah berkembang menjadi bentuk polimorfik, metamorfik, hingga oligomorfik. Dengan perkembangan massif ini, perangkat lunak antivirus konvensional tidak akan mampu mengatasinya dengan baik. Hal ini disebabkan oleh kemampuan malware untuk menyebarkan dirinya dengan pola sidik jari dan perilaku yang berbeda. Oleh karena itu, diperlukan antivirus cerdas berbasis machine learning yang mampu mendeteksi malware berdasarkan perilaku bukan sidik jari. Penelitian ini berfokus pada implementasi model machine learning dalam deteksi malware dengan menggunakan algoritma ensemble dan seleksi fitur untuk mencapai kinerja yang baik. Algoritma ensemble yang digunakan adalah Random Forest, dievaluasi dan dibandingkan dengan k-Nearest Neighbor dan Decision Tree sebagai state-of-the-art. Untuk meningkatkan kinerja klasifikasi dalam hal kecepatan proses, metode seleksi fitur yang diterapkan adalah Information Gain dengan 22 fitur. Hasil tertinggi dicapai dengan menggunakan algoritma Random Forest dan metode seleksi fitur Information Gain, mencapai skor 99.0% untuk akurasi dan F1-Score. Dengan mengurangi jumlah fitur, kecepatan pemrosesan dapat ditingkatkan hingga hampir 5 kali lipat.
Model Hybrid Random Forest dan Information Gain untuk meningkatkan Performa Algoritma Machine Learning pada Deteksi Malicious Software Rafrastara, Fauzi Adi; Ghozi, Wildanil; Sani, Ramadhan Rakhmat; Handoko, L. Budi
Jurnal Informatika dan Rekayasa Perangkat Lunak Vol. 6 No. 2 (2024): September
Publisher : Universitas Wahid Hasyim

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The evolution of malware, or malicious software, has raised increasing concerns, targeting not only computers but also other devices like smartphones. Malware is no longer just monomorphic but has evolved into polymorphic, metamorphic, and oligomorphic forms. With this massive development, conventional antivirus software is becoming less effective at countering it. This is due to malware's ability to propagate itself using different fingerprint and behavioral patterns. Therefore, an intelligent machine learning-based antivirus is needed, capable of detecting malware based on behavior rather than fingerprints. This research focuses on the implementation of a machine learning model for malware detection using ensemble algorithms and feature selection to achieve optimal performance. The ensemble algorithm used is Random Forest, evaluated and compared with k-Nearest Neighbor and Decision Tree as state-of-the-art methods. To enhance classification performance in terms of processing speed, the feature selection method applied is Information Gain, with 22 features. The highest results were achieved using the Random Forest algorithm and Information Gain feature selection method, reaching a score of 99.0% for accuracy and F1-Score. By reducing the number of features, processing speed can be increased by almost fivefold.
The Implementation of AWS Cloud Technology to Enhance the Performance and Security of the Pharmacy Cashier Management System Hendy Kurniawan; L. Budi Handoko; Valentino Aldo
INOVTEK Polbeng - Seri Informatika Vol. 10 No. 1 (2025): Maret
Publisher : P3M Politeknik Negeri Bengkalis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35314/x0rctv54

Abstract

 This study examines the implementation of Amazon Web Services (AWS) in the MEKATEK pharmacy cashier management system to address the limitations of traditional systems, such as slow transaction processing, data loss risks, and challenges in handling transaction surges. The prototyping method was employed, involving user requirements analysis through interviews and observations, followed by iterative development of core features like inventory management, transactions, reporting, and data backups. Black box testing demonstrated a 100% success rate for core functionalities. Performance analysis recorded stable CPU utilisation below 5% under normal workloads and the ability to handle throughput up to 2532 packets/minute. System optimisation reduced AWS operational costs to IDR 150,000–160,000 per month. AWS implementation improved operational efficiency, strengthened data security through encryption and role-based access control, and minimised human errors. Initial user feedback indicated faster workflows, although adjustments are needed for users with limited technical backgrounds. This study recommends further development, including AI-based analytics and digital payment integration, to enhance MEKATEK’s functionality and competitiveness in the future.
Optimasi Analisis Sentimen Lowongan Kerja di Twitter Dengan XGBoost-Vader dan Evaluasi SMOTE Borderline Ja'far, Luthfi; Handoko, L. Budi
Jurnal Pendidikan dan Teknologi Indonesia Vol 5 No 4 (2025): JPTI - April 2025
Publisher : CV Infinite Corporation

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52436/1.jpti.741

Abstract

Perkembangan komunikasi digital telah menjadikan Twitter sebagai platform utama dalam rekrutmen di Indonesia. Namun, analisis sentimen pada platform ini masih jarang diterapkan secara optimal, padahal dapat memberikan wawasan penting bagi pencari kerja dan perekrut dalam memahami persepsi publik terhadap lowongan kerja. Penelitian ini mengembangkan model analisis sentimen menggunakan XGBoost dan VADER untuk mengklasifikasikan postingan lowongan kerja berbahasa Indonesia ke dalam tiga kategori: positif, negatif, dan netral. Dataset terdiri dari 2.181 postingan, dengan rincian 1.711 netral, 414 positif, dan 56 negatif. Untuk menangani ketidakseimbangan data, diterapkan Synthetic Minority Over-sampling Technique (SMOTE) Borderline, yaitu teknik penyeimbangan data yang secara selektif menghasilkan sampel sintetis pada batas keputusan. Namun, teknik ini tidak meningkatkan akurasi model secara signifikan. Sebelum tuning, akurasi model konsisten di 99,95% hingga 100%, sementara setelah tuning, akurasi awalnya sedikit lebih rendah tetapi kemudian stabil di 100%. Evaluasi menggunakan classification report, confusion matrix, dan Stratified K-Fold Cross Validation menunjukkan bahwa model tetap stabil dan mampu menggeneralisasi data dengan baik tanpa indikasi overfitting. Dibandingkan pendekatan sebelumnya, penelitian ini menunjukkan bahwa kombinasi XGBoost dan VADER tanpa balancing data tambahan tetap mampu memberikan analisis sentimen yang lebih akurat dan stabil untuk platform lowongan kerja di Indonesia. Hasil ini berkontribusi dalam pengembangan model analisis sentimen berbasis machine learning yang lebih sesuai dengan karakteristik bahasa Indonesia, serta membuka peluang penelitian lebih lanjut dalam analisis opini publik di media sosial.
Deteksi dan Pencegahan Web Defacing Judi Online dengan Wazuh SIEM dan Snort IDS Berbasis Signature Reza Pahlevi, Mohammad Rizky; Umam, Chaerul; Handoko, L. Budi
Jurnal Algoritma Vol 22 No 1 (2025): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.22-1.2220

Abstract

Web defacing attacks, where websites are replaced with unwanted content, such as online gambling advertisements, pose a serious threat to the integrity and reputation of websites, especially those belonging to government agencies. This research aims to detect and prevent web defacing attacks containing online gambling content by combining Wazuh Security Information and Event Management (SIEM) and Snort signature-based Intrusion Detection System (IDS). Wazuh is used to monitor and collect activity logs in real-time when suspicious activity is detected. Meanwhile, Snort IDS acts as a signature-based intrusion detection system that can recognize web defacing attack patterns through predefined rules for online gambling content. This research was conducted by building a web defacing attack simulation environment on the server, then testing the response and effectiveness of Wazuh and Snort in detecting and preventing attacks. The test results show that the combination of Wazuh SIEM and Snort IDS can detect and prevent web defacing attacks with a very high accuracy rate, namely 100% of attacks can be detected by Wazuh File Integrity Monitoring and 76% for Snort IDS. The implementation of this system is expected to help improve website security, especially those managed by public institutions, from web defacing threats.
Development of a Website-Based Facilities and Infrastructure Rental System using the Rapid Application Development Method Valentino Aldo; L. Budi Handoko
INOVTEK Polbeng - Seri Informatika Vol. 10 No. 2 (2025): Juli
Publisher : P3M Politeknik Negeri Bengkalis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35314/wcqyg231

Abstract

To improve the efficiency and transparency of the management of facilities and infrastructure at the Semarang City Youth and Sports Office, a web-based rental system was developed with the RAD approach. Evaluation using the time measurement technique showed that the booking process time was reduced from 10 minutes to 3 minutes, and payment validation, which previously took up to 1 hour, now takes place automatically in seconds. The system was built using Express.js based on Node.js for an efficient and structured backend, React.js for an interactive and responsive frontend, and MySQL as the main database. The system design uses visual aids such as use case diagrams, activity diagrams, and entity relationship diagrams. Testing was carried out using black box testing using the equivalence partitioning technique. As a result, the system meets all functional requirements and increases operational efficiency by up to 70% through payment gateway integration. Further development, it is recommended to add reporting and analysis features to support decision making.
Vulnerability Analysis on Semarang City Road Section Information System Website Using VAPT Method Hanif Setia Nusantara; L. Budi Handoko; Maulana Ikhsan; Chaerul Umam
INOVTEK Polbeng - Seri Informatika Vol. 10 No. 2 (2025): Juli
Publisher : P3M Politeknik Negeri Bengkalis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35314/gdaky847

Abstract

Web-based public service applications in the digital governance era are increasingly vulnerable to cyber threats. This study analyzes the vulnerability of the Semarang City Road Information System website quantitatively using the Vulnerability Assessment and Penetration Testing (VAPT) method to evaluate its effectiveness in identifying security gaps. This system is part of an e-government service providing road infrastructure information but, like other technology-based systems, is susceptible to exploitation. The VAPT method used includes two main stages: Vulnerability Assessment to identify weaknesses and Penetration Testing to simulate attacks. The study identified 5 potential vulnerabilities: SQL Injection, Credit Card Number Disclosure, Insecure Direct Object Reference (IDOR), Cross-Site Scripting (XSS), and Error Message on Page. However, 80% of these were false positives, effectively filtered by Alibaba Cloud’s Web Application Firewall (WAF). The IDOR vulnerability was confirmed as valid, allowing unauthorized access to sensitive data through manipulation of the ID parameter in the URL. The original contribution of this research is the specific recommendation for implementing Indirect Object References mechanisms such as ID encryption, as well as emphasizing the need for comprehensive routine testing to improve security and prevent potential data misuse.
DIGITAL SIGNATURE PADA CITRA MENGGUNAKAN RSA DAN VIGENERE CIPHER BERBASIS MD5 Handoko, Lekso Budi; Umam, Chaerul; Setiadi, De Rosal Ignatius Moses; Rachmawanto, Eko Hari
Simetris: Jurnal Teknik Mesin, Elektro dan Ilmu Komputer Vol 10, No 1 (2019): JURNAL SIMETRIS VOLUME 10 NO 1 TAHUN 2019
Publisher : Fakultas Teknik Universitas Muria Kudus

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24176/simet.v10i1.2212

Abstract

Salah satu teknik yang populer untuk mengamankan data dengan tingkat keamanan yang tinggi yaitu kriptografi. Berbagai penelitian telah dilakukan dengan menggabungkan kunci simteris dan kunci asimteris untuk mendapatkan keamanan ganda. Dalam makalah ini, tanda tangan digital diterapkan melalui Rivest Shamir Adleman (RSA) sebagai algoritma kunci asimteris yang akan digabung dengan algoritma kunci simteris Vigenere Cipher. RSA yang tahan terhadap serangan karena menggunakan proses eksponensial dan kuadrat besar dapat menutupi kelemahan Vigenere Cipher, sedangkan Vigenere Cipher dapat mencegah kemunculan huruf yang sama dalam cipher yang mempunyai pola tertentu. Vigenere cipher mudah diimplementasikan dan menggunakan operasi substitusi. Untuk mengkompresi nilai numerik yang dihasilkan secara acak, digunakan fungsi hash yaitu Message Digest 5 (MD5). percobaan dalam makalah ini telah memberikan kontribusi dalam peningkatan kualitas enkripsi dimana citra digital dioperasikan dengan MD5 yang kemudian hasilnya akan diubah menjadi RSA. Fungsi hash awal yaitu 32 karakter diubah menjadi 16 karakter yang akan menjadi inputan untuk proses RSA dan Vigenere Cipher. Pada citra berwarna yang digunakan sebagai media operasi, akan dilakukan pengecekan apakah citra tersebut sudah melalui proses digital signature
Implementation Of Extreme Gradient Boosting Algorithm For Predicting The Red Onion Prices Saputri, Pungky Nabella; Alzami, Farrikh; Saputra, Filmada Ocky; Andono, Pulung Nurtantio; Megantara, Rama Aria; Handoko, L Budi; Umam, Chaerul; Wahyudi, Firman
Moneter: Jurnal Keuangan dan Perbankan Vol. 11 No. 1 (2023): APRIL
Publisher : Universitas Ibn Khladun Bogor

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (656.456 KB) | DOI: 10.32832/moneter.v11i1.55

Abstract

Red Onion or the Latin name Allium Cepa is included in the group of vegetable plants that are needed by the public for food needs. Red Onions are one of the seasonal crops so their availability can change in the market which causes price instability due to a lack of supply of production by several factors: 1) not yet it's harvest time, 2) crop attacked disease pests and fungi, and 3) weather factor. Therefore, a study is needed to predict red onion prices, so that it can be used as information for the government to stabilize red onion prices. The method used in this study is CRISP-DM and the Extreme Gradient Boosting algorithm to predict the price of red onions by taking data samples from Tegal and Pati Cities. The results of this study are that the Extreme Gradient Boosting algorithm is able to produce Tegal District Root Mean Square Error (RMSE) values of 5107.97% and Mean Absolute Percentage Error (MAPE) values of 0.17%. For prediction results with Pati Regency data samples, it produces a Root Mean Square Error (RMSE) value of 6049.74% and a Mean Absolute Percentage Error (MAPE) of 0.17%.
PREDIKSI EMAIL PHISING MENGGUNAKAN SUPPORT VECTOR MACHINE Umam, Chaerul; Handoko, L. Budi
Semnas Ristek (Seminar Nasional Riset dan Inovasi Teknologi) Vol 8, No 01 (2024): SEMNAS RISTEK 2024
Publisher : Universitas Indraprasta PGRI

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30998/semnasristek.v8i01.7138

Abstract

Email phising merupakan salah satu bentuk kejahatan di internet yang dapat merugikan banyak orang. Ketika seseorang sudah terkena phising maka data data orang tersebut dapat hilang dan digunakan oleh orang yang tidak bertanggung jawab. Pada penelitian ini, akan melakukan proses klasiifkasi email phising dengan menggunakan bantuan machine learning yaitu algoritma SVM. Dataset yang digunakan pada penelitian ini yaiitu merupakan dataset yang berisi body email yang terdiri dari total 18650 data yang terdiri dari 11322 data safe email dan 7328 data phising email. Dari data tersebut, akan dibagi menjadi 70% data pelatihan dan 30% data pengujian. Setelah dilakukan proses pengujian pada penelitian ini, algoritma SVM yang digunakan mendapatkan akurasi pengujian sebesar 84.56%.