Claim Missing Document
Check
Articles

Comparison of Dimensionality Reduction Techniques to Improve Performance and Efficiency of Logistic Regression in Network Anomaly Detection Ahfa, Mokhamad Isna Marzuki; Hakim, Lukman; Rosadi, Muhammad Imron
Journal of Information Technology and Cyber Security Vol. 3 No. 1 (2025): January
Publisher : Department of Information Systems and Technology, Faculty of Intelligent Electrical and Informatics Technology, Universitas 17 Agustus 1945 Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30996/jitcs.12212

Abstract

Network anomaly detection is a crucial process to identify abnormal network traffic, which may pose a security threat. This research aims to improve the performance and efficiency of Logistic Regression (LR) in network anomaly detection by applying dimension reduction techniques, such as Principal Component Analysis (PCA), Truncated Singular Value Decomposition (TSVD), t-Distributed Stochastic Neighbor Embedding (t-SNE), and Independent Component Analysis (ICA). The performance of each dimension reduction method is evaluated based on accuracy, precision, recall, F1-score, and computation time. The results show that TSVD provides the best performance with 95.86% accuracy, 0.96 precision, 0.96 recall, 0.95 F1-score, and 13.83 seconds computation time. In contrast, ICA showed the worst performance, especially in precision, recall, and F1-score, with values of 0.73, 0.83, and 0.78, respectively. Meanwhile, although t-SNE produces competitive accuracy, it has a high computational cost with an execution time of 1698.54 seconds. These findings show that choosing the right dimension reduction algorithm not only improves detection performance but also supports data processing efficiency, making it highly relevant for large-scale network security scenarios. Keywords: dimensionality reduction, Logistic Regression, network anamoly detection, performance evaluation, Truncated Singular Value Decomposition.
Comparison of Dimensionality Reduction Techniques to Improve Performance and Efficiency of Logistic Regression in Network Anomaly Detection Ahfa, Mokhamad Isna Marzuki; Hakim, Lukman; Rosadi, Muhammad Imron
Journal of Information Technology and Cyber Security Vol. 3 No. 1 (2025): January
Publisher : Department of Information Systems and Technology, Faculty of Intelligent Electrical and Informatics Technology, Universitas 17 Agustus 1945 Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30996/jitcs.12212

Abstract

Network anomaly detection is a crucial process to identify abnormal network traffic, which may pose a security threat. This research aims to improve the performance and efficiency of Logistic Regression (LR) in network anomaly detection by applying dimension reduction techniques, such as Principal Component Analysis (PCA), Truncated Singular Value Decomposition (TSVD), t-Distributed Stochastic Neighbor Embedding (t-SNE), and Independent Component Analysis (ICA). The performance of each dimension reduction method is evaluated based on accuracy, precision, recall, F1-score, and computation time. The results show that TSVD provides the best performance with 95.86% accuracy, 0.96 precision, 0.96 recall, 0.95 F1-score, and 13.83 seconds computation time. In contrast, ICA showed the worst performance, especially in precision, recall, and F1-score, with values of 0.73, 0.83, and 0.78, respectively. Meanwhile, although t-SNE produces competitive accuracy, it has a high computational cost with an execution time of 1698.54 seconds. These findings show that choosing the right dimension reduction algorithm not only improves detection performance but also supports data processing efficiency, making it highly relevant for large-scale network security scenarios. Keywords: dimensionality reduction, Logistic Regression, network anamoly detection, performance evaluation, Truncated Singular Value Decomposition.
Pelatihan Dasar Fotografi untuk Mendukung Digitalisasi Produk Kopi Bubuk di UMKM Linusco Kabupaten Pasuruan Ernawati; Ainiyah, Roisatul; Palupi, Hapsari Titi; Syarwani, Muhammad; Hasyim, Mochamad; Huda, Miftahul; Ridha, Faishal Ananta; Ahwan, Zainul; Hakim, Lukman; Sulhan, Muhammad; Swasono, Muh Aniar Hari
DEDIKASI SAINTEK Jurnal Pengabdian Masyarakat Vol. 3 No. 3 (2024): Desember 2024
Publisher : Al-Hijrah Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58545/djpm.v3i3.439

Abstract

Platform digital menjadi salah satu media utama dalam memperluas jangkauan pemasaran produk, terutama bagi Usaha Mikro, Kecil, dan Menengah (UMKM). Kualitas foto produk memiliki peran yang sangat penting untuk meningkatkan daya tarik visual dan memperluas jangkauan pemasaran produk. Tujuan dari kegiatan ini adalah untuk meningkatkan kapasitas UMKM Linusco dalam memanfaatkan fotografi produk sebagai strategi pemasaran digital. Metode pelaksanaan meliputi tiga tahapan utama: perencanaan, pelaksanaan, dan evaluasi. Tahap perencanaan mencakup identifikasi kebutuhan, penyusunan materi, dan penyediaan alat pendukung, sedangkan tahap pelaksanaan difokuskan kegiatan pelatihan. Hasil pelatihan menunjukkan peningkatan signifikan dalam pemahaman konsep fotografi produk, kemampuan pengambilan foto, dan penguasaan aplikasi pengeditan, sebagaimana ditunjukkan melalui evaluasi hasil praktik peserta. Peserta berhasil menghasilkan foto produk yang lebih menarik dan profesional, siap digunakan untuk promosi digital. Pendekatan kegiatan pelatihan mampu meningkatkan pengetahuan dan keterampilan peserta.
Implementation of Integrated Smart System Platform in Improving the Quality of Public Services through Smart Village In Pucangsari Village, Pasuruan Regency Hakim, Lukman; Rosadi, Muhammad Imron; Prianto, Agus

Publisher : LPPM Universitas Yudharta Pasuruan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35891/js.v6i1.4499

Abstract

The purpose of this service is to increase the knowledge and skills of the Pucangsari village government in the concept and implementation of a smart village based on an integrated smart system platform through village administration governance, reporting, and utilization of public service applications. The method used in the integrated smart system platform development program in village government services includes three stages; the first stage of preparation by conducting surveys and interviews related to the potential and problems faced by prospective program partners; the second stage of making and developing Integrated smart system Platform Applications; and the third stage of capacity building in the form of workshops, training and socialization. The results of this community service are increased knowledge related to smart villages, the ability to use integrated smart system platform applications and improved public services which include correspondence, billing and payment of land and building taxes, as well as the existence of a community aspiration space as an improvement in the Pucangsari village government performance system.
Assistance of Tourism Cyber Public Relations in Increasing Publicity of Virtual Accessibility through Integrated Tourism Smart Platform (ITSP) in Wonokitri Tourism Village Pasuruan Hakim, Lukman; Ahwan, Zainul; Rosadi, Muhammad Imron; Sanjaya, Cahya Bagus

Publisher : LPPM Universitas Yudharta Pasuruan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35891/js.v7i1.5531

Abstract

The hinterland area of Mount Bromo possesses exotic natural and cultural potential that can be developed into a leading tourist destination. However, the management and promotion of this potential face challenges, particularly in terms of tourism communication skills and the utilization of technology. The activity is to enhance the capacity of local tourism actors by mastering communication skills based on Tourism Cyber Public Relations and improving their ability to utilize technology to support more innovative and sustainable tourism services. Additionally, this program aims to facilitate the development of an integrated digital identity for the tourist village. The assistance program was carried out through five stages: socialization, training, technology implementation, mentoring, and sustainability evaluation. The approach employed was the Community Organizer and Community Development methods, with the main subject being the Raga Wulan Tourism Awareness Group in Wonokitri Village. The results of the activities showed a 25% increase in the knowledge of tourism actors based on pre-test and post-test results conducted during the program. Furthermore, this initiative successfully enhanced the tourism actors' skills in tourism communication, particularly in the field of Tourism Cyber Public Relations. It also resulted in the creation of branding and the development of social media and the official website of Edelweiss Wonokitri Tourism Village, as well as the realization of the Integrated Tourism Smart Platform.
Characteristics of ASTM A36 steel mechanical properties in pack carburizing with carburizing agent coconut shell charcoal and goat bone powder mixed Sujita, S.; Okariawan, I.D.K.; Hakim, L.
Dinamika Teknik Mesin: Jurnal Keilmuan dan Terapan Teknik Mesin Vol 13, No 1 (2023): Dinamika Teknik Mesin: Jurnal Keilmuan dan Terapan Teknik Mesin
Publisher : Universitas Mataram

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29303/dtm.v13i1.619

Abstract

In the study the specimens were ASTM A36 low carbon steel pack  carburizing at 950℃, soaking time variations of 1, 2 and 3 hours. The carburizing agent is a mixture of coconut shell charcoal and goat bone powder (Capra aegagrus hircus). The size of the powder is 100 mesh, with variations in weight percentage between goat bone powder and coconut shell charcoal: 10%:90%, 20% : 80%, and 30% : 70%. Furthermore, the impact test, hardness, microstructure test and composition test were carried out, to determine the mechanical properties of the specimens. The results of the study show, soaking time and the composition of the carburizing agent affect the mechanical properties of the specimen The lowest impact toughness of 0,0576 joules/mm2 was obtained by the specimen at a soaking  time of 1 hour, the composition of the carburizing agent  (% wt goat bone powder: coconut shell charcoal) 20% : 80%. The highest surface hardness value was 854.734 Kg/mm2 at 3 hours soaking  time with 10%:90% carburizing media composition. The results of microstructure observations showed that in specimens treated with pack carburizing ferrite structure, pearlite turned into martensite.
ARSITEKTUR ENSEMBLE U-NET UNTUK SEGMENTASI KANKER PAYUDARA OTOMATIS PADA GAMBAR MAMMOGRAM Sanjaya, Cahya Bagus; Lutfi, Moch.; Hakim, Lukman; Rosadi, Muhammad Imron
NJCA (Nusantara Journal of Computers and Its Applications) Vol 9, No 2 (2024): Edisi Desember 2024
Publisher : Computer Society of Nahdlatul Ulama (CSNU) Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36564/njca.v9i2.378

Abstract

Kanker payudara masih menjadi salah satu penyebab utama kematian akibat kanker pada wanita di seluruh dunia. Deteksi dini melalui skrining rutin menggunakan mammogram terbukti efektif dalam mengurangi angka kematian. Namun, interpretasi mammogram secara manual memerlukan waktu, bersifat subjektif, dan sering kali membutuhkan radiolog yang berpengalaman. Untuk mengatasi tantangan ini, penelitian ini mengusulkan arsitektur Ensemble U-Net untuk melakukan segmentasi kanker payudara secara otomatis pada citra mammogram. Proses segmentasi melibatkan beberapa langkah, termasuk praproses (penghapusan latar belakang, penghapusan otot pektoral, peningkatan kontras, dan pengubahan ukuran), dilanjutkan dengan segmentasi menggunakan ensemble model: Inception V3-U-Net, ResNet50-U-Net, VGG19-U-Net, dan U-Net kustom. Segmentasi akhir dicapai dengan menggunakan voting soft dan filter Gaussian 2D untuk mereduksi noise, diikuti dengan thresholding untuk segmentasi biner. Pendekatan ensemble menunjukkan peningkatan akurasi segmentasi dengan menggabungkan kekuatan dari beberapa model U-Net. Kinerja model dievaluasi menggunakan metrik seperti akurasi, sensitivitas, spesifisitas, koefisien Dice, dan Intersection over Union (IoU). Hasil eksperimen menunjukkan bahwa Ensemble U-Net memiliki kinerja yang lebih baik dibandingkan dengan model individu, terutama pada citra mammogram yang kompleks.
Pendampingan Penerapan Standard Sanitation Operating Procedure (SSOP) pada Proses Produksi Sari Murni Jeruk di UMKM Panda Sirup Kabupaten Pasuruan, Jawa Timur Ernawati, Ernawati; Ainiyah, Roisatul; Palupi, Hapsari Titi; Syarwani, Muhammad; Hasyim, Mochamad; Huda, Miftahul; Murtadlo, Kholid; Yusebla, Agil; As'idah, Fizriyatul; Ahwan, Zainul; Hakim, Lukman; Sulhan, Muhammad; Swasono, Muh. Aniar Hari
Jurnal Inovasi Pengabdian dan Pemberdayaan Masyarakat Vol 4 No 2 (2024): JIPPM - Desember 2024
Publisher : CV Firmos

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54082/jippm.669

Abstract

Panda Sirup merupakan mitra UMKM yang berada di Kecamatan Gondang Wetan Kabupaten Pasuruan dengan produk utama berupa sari murni jeruk nipis dan lemon. Permasalahan utama yang dihadapi oleh mitra adalah belum menerapkan jaminan keamanan pangan yang sesuai dengan standar dengan penerapan Standard Sanitation Operating Procedure (SSOP). Tujuan dari kegian pengabdian ini adalah untuk meningkatkan kualitas dan jaminan kemanan pangan produk melalui kegiatan pendampingan penerapan SSOP. Metode pelaksanaan yang digunakan terdiri dari kegiatan observasi dan wawancara, pelatihan, pendampingan, serta monitoring & evaluasi. Hasil dari kegiatan pendampingan ini memberikan dampak berupa meningkatnya pengetahuan dan pemahaman mitra dari 50% menjadi 85%. Pada aspek proses produksi pengetahuan dan pemahaman mitra meningkat dari 60% menjadi 90%. Pada aspek penerapan SSOP oleh mitra pengetahuan dan pemahaman mitra meningkat dari 55% menjadi 88%. Selain itu telah disusun dokumen SSOP dan pendampingan perbaikan fasilitas sesuai standar SSOP, serta beberapa saran teknis terkait metode pembersihan peralatan yang baik dan benar. Dengan penerapan SSOP yang baik dan benar diharapkan memberikan dampak terhadap peningkatan mutu dan keamanan pangan produk sari murni jeruk di UMKM Panda Sirup.
Monte Carlo method at the 24 game and its application for mathematics education Fitrianawati, Meita; Aliansyah, Zulhaj; Peni, Nur Robiah Nofikusumawati; Farid, Imam Wahyudi; Hakim, Lukman
Journal of Honai Math Vol. 5 No. 2 (2022): Journal of Honai Math
Publisher : Universitas Papua

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30862/jhm.v5i2.250

Abstract

Students often find mathematics a challenging subject and turn it into a scourge for them. Game-based learning, such as “24-card game”, help engage students in a self-paced and fun learning process and thus may overcome students’ math anxiety and promote mental math skills. This research aims to examine how the 24-card game works using the Monte Carlo method and the possibility to overcome students' mathematics anxiety. The meta-analysis method was used to explain Monte Carlo’s simulation to solve the solution for all possible combinations of cards in the game and respectively assign difficulty levels. The student's proficiency level was evaluated based on the divergence value in the number of guesses required to solve the dealt combination at 87% to show full proficiency. The evaluation could also show the math difficulty of advanced operations, such as fractions and grouping games. This game is more efficient in developing students' mental math skills compared to a conventional and rigidly structured classroom lecture.
KLASIFIKASI SERANGAN DDOS MENGGUNAKAN REQURSIVE FEATURE ELIMINATION DAN GRADIENT BOOSTING Candra Adi Lesmana; Lukman Hakim
Jurnal Informatika dan Rekayasa Elektronik Vol. 8 No. 1 (2025): JIRE APRIL 2025
Publisher : LPPM STMIK Lombok

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36595/jire.v8i1.1396

Abstract

Keamanan internet menjadi tantangan penting seiring dengan pertumbuhan layanan teknologi informasi yang semakin pesat. Serangan Distributed Denial of Service (DDoS) merupakan salah satu ancaman serius yang dapat menyebabkan crash pada server dan sistem jaringan dengan cara membanjiri jaringan dengan paket atau permintaan yang berlebihan. Dalam penelitian ini, diterapkan metode seleksi fitur Recursive Feature Elimination (RFE) pada proses klasifikasi serangan DDoS menggunakan algoritma Gradient Boosting. Tujuan Penelitian ini untuk melakukan penerapan algoritma seleksi fitur Reqursive Feature Elimination (RFE) untuk mengurangi jumlah fitur yang ada dalam data. Hasil pengujian menunjukkan bahwa penggunaan Gradient Boosting dengan parameter terbaik menghasilkan kinerja yang sangat baik, dengan akurasi sebesar 99,9%, presisi 99,8%, recall 99,9%, dan nilai F1 sebesar 99,9%. Kombinasi metode Reqursive Feature Elimination (RFE) dengan Gradient Boosting, di mana 10 fitur terbaik dipilih, tidak mempengaruhi hasil kinerja model secara signifikan, tetapi penerapan seleksi fitur ini berhasil mengurangi waktu komputasi secara signifikan dari 5463,24 detik menjadi 1023,19 detik tanpa mengorbankan kinerja model. Hasil ini menunjukkan bahwa metode seleksi fitur yang tepat sangat penting dalam meningkatkan efisiensi komputasi tanpa mengorbankan performa model dalam deteksi serangan DDoS.