Publikasi Universitas Mercu Buana
Not a member yet
7303 research outputs found
Sort by
Implementasi Layanan Broadband Network Gateway Dengan Mengoptimalkan Perangkat Metro Ethernet Menggunakan Metode BGP, VPRN, dan Subscriber Management (BVSM)
No full textPenelitian ini bertujuan untuk mengeksplorasi potensi perangkat router metro ethernet (Nokia 7750 SR-7) dalam menyediakan layanan Broadband Network Gateway (BNG) dengan mengimplementasikan metode BGP, VPRN, dan Subscriber Management (BVSM). Metode BVSM menunjukkan keberhasilan perangkat metro ethernet dalam pengelolaan subscribers BNG dengan transisi yang efisien dari penggunaan service layer 2 (VPLS) ke service layer 3 (VPRN). Konfigurasi parameter tambahan yang diperlukan pada service VPRN, seperti policy-option, local DHCP server, dan subscriber interface memastikan optimalitas layanan BNG pada perangkat metro-ethernet. Evaluasi potensi keuntungan dari implementasi layanan BNG pada perangkat metro-ethernet menunjukkan peningkatan pada efisiensi jaringan dan pengurangan titik kegagalan dengan tidak perlu lagi digunakannya perangkat Broadband Remote Access Server (BRAS) untuk mengelola layanan BNG. Dengan manajemen layanan BNG yang lebih terpusat di perangkat metro ethernet, proses operasional dan maintenance dapat dilakukan dengan efektif, meningkatkan kualitas layanan dan manajemen pelanggan secara menyeluruh dengan cara mengoptimalkan alokasi IP Pool, monitoring user session, dan mengatasi gangguan layanan dengan lebih cepat
IMPLEMENTATION OF MARKETING PUBLIC RELATIONS STRATEGIES IN INCREASING THE EXISTENCE OF SAYURBOX
No full textIn the era of global competition, companies are required to enhance performance and business activities. Marketing plays a crucial role, yet extensive promotions incur significant costs. The internet and online shopping have transformed businesses, particularly e-grocery. Sayurbox leverages digital platforms but faces challenges like changing consumer behavior and layoffs due to mismatched target markets. Companies must build a strong reputation beyond product quality. MPR acts as a bridge between companies and consumers. This study analyzes MPR implementation in enhancing Sayurbox's existence. The research methodology employs a qualitative approach with a case study. Data was collected through interviews with the public relations manager, and five consumers of Sayurbox. The implementation of Marketing Public Relations (MPR) by Sayurbox includes structured PR strategies, a focus on positive image building through company values, and transparent crisis management. Effective communication strategies are used for new product launches. Sayurbox adopts a combination of pull, push, and pass strategies to expand market reach. Pull strategies emphasize customer interaction via social media and events. Push strategies include promotions, bundle deals, loyalty programs, and endorsements. Pass strategies involve community relations and social responsibility. Strong negotiation skills are also a key advantage
Quality improvement through 8D methodology: an automotive industry case study
No full textNowadays quality is a key factor for the success of a product or company to survive and be accepted by consumers. Repeated quality problem in automotive industry become one of factor that influence company performance. However the problem is visual aspect, if this repetitive will increase customer dissatisfaction. Problem covers pillar loop scratch becomes the repeated and worst problem in Feb–March 2020. This research aims to assist in quality improvement by using the eight discipline (8D) method to solve the problem “cover pillar loop scratch. An 8D method is a problem-solving tool that has complete and systematic corrective action stages in problem-solving, which is for short-term action and long-term corrective action also effective in preventing the re-occurring problem. Result the implementation of the 8D methodology in problem-solving seat belt production problems in the SB3 line of Automotive Industry succeeded in reducing and eliminating the number of problems “cover pillar loop scratch” from 60 pcs in March 2020 to 0 in July 2020
Implementasi Data Mining dan Machine Learning untuk Segmentasi Pelanggan: Pendekatan Hybrid Menggunakan Big Data
No full textDeteksi dini penyakit jantung merupakan langkah penting untuk meningkatkan kualitas diagnosis dan perawatan pasien. Namun, metode prediksi manual yang sering digunakan tenaga medis memiliki keterbatasan dalam efisiensi waktu, akurasi, dan kemampuan menangani volume data yang besar. Dalam bidang kecerdasan buatan, algoritma machine learning seperti Adaptive Boosting (AdaBoost), Gradient Boosting, dan Extreme Gradient Boosting (XGBoost) menawarkan potensi untuk meningkatkan akurasi prediksi, terutama dalam mengatasi tantangan pada dataset kecil yang sering mengalami ketidakseimbangan kelas dan risiko overfitting. Penelitian ini bertujuan untuk menganalisis kinerja ketiga algoritma boosting tersebut dalam memprediksi penyakit jantung. Hasil penelitian menunjukkan bahwa XGBoost memberikan performa terbaik dengan akurasi sebesar 84.78% dan ROC-AUC 0.9410, menjadikannya algoritma paling efektif dalam menangani pola data yang kompleks. Gradient Boosting menjadi model paling efisien dengan waktu pelatihan tercepat, yaitu 0.3655 detik, dengan akurasi dan ROC-AUC yang kompetitif. Sementara itu, AdaBoost menunjukkan kelemahan dalam menangani ketidakseimbangan kelas tetapi tetap memberikan hasil yang baik untuk kelas mayoritas. Berdasarkan evaluasi precision, recall, dan F1-score, XGBoost direkomendasikan untuk aplikasi prediksi penyakit jantung, terutama dalam situasi yang memerlukan akurasi tinggi, sedangkan Gradient Boosting cocok untuk kebutuhan real-time
Klasifikasi Sentimen iPhone Bekas di Tokopedia menggunakan Naïve Bayes dan Support Vector Machine
No full textKenaikan harga iPhone baru mendorong meningkatnya pembelian iPhone second di platform e-commerce seperti Tokopedia. Namun, konsumen masih menghadapi berbagai risiko terkait kondisi perangkat, performa komponen, dan keaslian yang umumnya teridentifikasi melalui ulasan pengguna. Penelitian ini bertujuan menganalisis sentimen dari 1.863 ulasan iPhone second untuk memperoleh gambaran objektif mengenai pengalaman konsumen. Teks ulasan diproses menggunakan TF-IDF sebagai representasi fitur dan SMOTE untuk mengatasi ketidakseimbangan kelas. Dua algoritma Naive Bayes dan Support Vector Machine (SVM) dibandingkan untuk menilai efektivitas klasifikasi. Hasil pengujian menunjukkan bahwa SVM memberikan performa terbaik dengan akurasi 96%, melampaui Naive Bayes yang mencapai 93%. Analisis lebih lanjut menemukan bahwa ulasan positif umumnya berkaitan dengan kualitas fisik dan kecepatan pengiriman, sedangkan ulasan negatif banyak menyoroti isu teknis serta keaslian perangkat. Penelitian ini berkontribusi pada penguatan literatur analisis sentimen e-commerce melalui evaluasi komprehensif terhadap kombinasi TF-IDF + SMOTE serta perbandingan performa Naive Bayes dan SVM dalam klasifikasi opini konsumen. Temuan ini menyediakan dasar empiris untuk penelitian lanjutan mengenai penilaian kualitas produk bekas berbasis ulasan daring
Effect of Coconut Fiber and Coconut Shell Charcoal Composition on the Properties of PVC-Reinforced Composite Brake Pads
No full textThe increasing concern over the health hazards associated with asbestos-based brake pads has driven the development of eco-friendly alternatives using natural fiber-reinforced composites. This study aims to fabricate and evaluate a sustainable brake pad material using coconut fiber as reinforcement, coconut shell charcoal powder as filler, and polyvinyl chloride (PVC) as the matrix. The composite was manufactured using the hot press method at a temperature of 180°C and a pressure of 7 MPa, conditions selected to optimize resin curing and interfacial bonding. A key focus of this research was to investigate the effect of solvent volume (cyclohexanone) used in the PVC resin preparation on the mechanical properties of the resulting composites. Three composite formulations were prepared with a constant composition of 70% coconut fiber, 5% charcoal powder, and 25% PVC resin, but with varying amounts of cyclohexanone solvent (200 mL, 150 mL, and 100 mL). The results revealed that reducing solvent content led to higher resin viscosity, which improved matrix–fiber bonding and increased both tensile strength and surface hardness. The optimal formulation—PVC Resin 3 with 100 mL of solvent—achieved a maximum tensile strength of 7.7 MPa and Shore D hardness of 72.2 HD, both of which meet the SAE J661-1997 standards for brake pad materials. This study confirms that solvent content is a critical factor influencing the density, strength, and durability of the composite. The findings support the feasibility of utilizing coconut-based agricultural waste in producing environmentally friendly brake pads with adequate mechanical performance
Machine Learning System untuk Mendeteksi Gerakan Tubuh Menggunakan Library Mediapipe
No full textCommunication with people with hearing and speech disabilities is often challenging. Sign language is the primary tool that helps them convey thoughts and feelings, but it is often difficult for those who are not used to it to understand. This project aims to develop a machine learning model to recognize hand gestures in spelling fingers using American Sign Language (ASL). The model uses image data and Computer Vision techniques to train a deep learning algorithm that can recognize signals in real-time through a camera. The system utilizes deep neural networks that work through layers of nodes to process, classify, and predict cues accurately
KETERLIBATAN KOMUNITAS INTI PADA PELESTARIAN KAMPUNG KETANDAN YOGYAKARTA
No full textPusaka bukan hanya dimaknai sebagai benda dan tak benda tetapi meluas pada kehidupan manusia didalamnya, terutama pada kawasan pusaka yang merupakan kawasan hunian yang dihuni secara turun temurun. Gerakan pelestarian berbasis masyarakat/komunitas menjadikan komunitas sebagai inti dari pelestarian berdampingan dengan site pusaka. Kampung Ketandan Yogyakarta merupakan salah satu kawasan pusaka di tengah kota Yogyakarta yang memiliki kekayaan asset benda, tak benda dan kehidupan. Kampung ketandan dikenal sebagai kampung Tionghoa pertama di Yogyakarta yang masih memiliki komunitas inti dan tinggal dalam kawasan secara turun temurun. Artikel ini bertujuan mengidentifikasi keterlibatan komunitas inti pada pelestarian Kampung Ketandan. Pengumpulan data dilakukan melalui eksplorasi lapangan melalui observasi, wawancara, dan penelusuran dokumen. Hasil data kemudian dianalisis berdasar 8 tangga tingkatan keterlibatan komunitas dalam pelestarian pusaka berdasarkan teori Arnstein (1969). Dari analisis yang dilakukan didapati bahwa peran komunitas inti masih pada tingkatan tokenisme sehingga masih perlu peningkatan untuk mencapai pelestarian yang berkelanjutan
Risk Analysis and Mitigation Strategies for the Supply Chain of Bread Products at MSME Muthia Bakery
Get PDFMuthia Bakery is a micro-scale business producing preservative-free bread and cakes in PPU Regency. This study aims to minimize supply chain risks by identifying, assessing, and developing mitigation strategies tailored to Muthia Bakery’s operations. Using the Supply Chain Operation Reference (SCOR) model, the study mapped key activities (plan, source, make, deliver, and return) to provide a clear structure for risk identification. This mapping facilitated a focused risk analysis using the Failure Mode Effect Analysis (FMEA) method, with Action Priority (AP) used to prioritize critical risks requiring immediate mitigation. Root causes were examined using a Fishbone Diagram, and AHP was applied to prioritize effective mitigation strategies. Results highlight five primary risks, including (1) inaccurate material purchase quantities (RPN 144), mitigated by improving data collection and analysis (weight 0.1000); (2) raw material returns (RPN 144), addressed through SOP development for quality control (weight 0.673); (3) material ordering delays (RPN 140), mitigated via inventory control (weight 0.635); (4) incorrect raw material quantities received (RPN 120), with double verification during ordering (weight 0.444); and (5) production scheduling errors (RPN 105), mitigated by improved time management (weight 0.701). This research provides a systematic risk management approach for micro-scale bakery supply chains, supporting continuity and efficient operational processes
Optimization of Flight Routes Employing the Simulated Annealing Method in the Context of the Indonesian National Airline Industry
Get PDFThis research was conducted in the context of significant air traffic growth in Indonesia, where the increasing number of passengers each year presents an opportunity for airlines to expand their route networks and reach more markets. To seize this opportunity, a national airline based in Jakarta conducted internal research to open new flight routes connecting several important cities in Indonesia, namely Jakarta, Kupang, Pangkal Pinang, Pekanbaru, Makassar, and Banjarmasin.This research aims to obtain an optimal flight route that can enhance the effectiveness and efficiency of the planned airline routes. The research utilizes the Simulated Annealing-Traveling Salesman Problem optimization method to achieve this objective. This method is employed to find the best solution in determining the shortest route that includes visits to each destination city.The initial proposed flight route by the airline was Banjarmasin-Pangkal Pinang-Pekanbaru-Jakarta-Kupang-Makassar, with a total distance of 5,127 km. However, the research yielded a different optimal flight route after conducting the optimization process using the Simulated Annealing-Traveling Salesman Problem method. The discovered optimal flight route is Pekanbaru-Pangkal Pinang-Jakarta-Banjarmasin-Makassar-Kupang, with a total distance of 3,256 km. A comparison between the initial and optimal routes reveals that the new route has a 36.49% shorter distance than the initial route