MEANS (Media Informasi Analisa dan Sistem)

Pengembangan Sistem Peramalan Permintaan Menggunakan Algoritma Support Vector Regression Untuk Optimalisasi Safety Stock Berbasis Web (Studi Kasus: JG Motor Sukabumi)

2025-01-15T03:16:00+01:00

This study aims to develop a web-based system utilizing Support Vector Regression (SVR) to predict motor vehicle spare part demand and optimize safety stock levels at JG Motor Sukabumi. The inventory management faces challenges such as fluctuating demand, supply delays, and overstock/stockout risks. To address these issues, SVR is chosen for its ability to handle non-linear and complex data, providing more accurate predictions than conventional methods. This research employs a descriptive quantitative approach with semi-experimental methods to test the SVR model's effectiveness and web-based system validity. The system features monthly demand prediction, safety stock calculation, historical data visualization, and interactive analytical reports. Development involves user requirement analysis, two-year historical sales data collection, data preprocessing, SVR model training with parameter optimization, and Flask-based integration. Black Box Testing ensures primary functions, such as input validation, prediction processing, and stock recommendation outputs, operate correctly. Results indicate the SVR model achieves high accuracy, reflected by low Mean Absolute Error (MAE) values. The web-based system is user-friendly for managers and operational staff to monitor demand and manage inventory efficiently. Moreover, the system supports strategic decision-making, enhancing JG Motor Sukabumi's operational efficiency and competitiveness in the automotive market.

Implementasi Metode Double Diamond Design Process pada UI/UX Aplikasi Manajemen Kesehatan dan Gizi Personal (WellFits)

2025-03-18T03:07:00+01:00

WellFits is an innovative digital solution for personal health and nutrition management, focusing on calorie deficit, stunting prevention in pregnant women, and nutrition consultation with experts. This study applies the Double Diamond Design Process to identify user needs and develop relevant features. The development process includes problem exploration, needs definition, solution development, and implementation with iterative eval_uation based on user feedback. Usability testing results showed an average score of 8.04 out of 10, indicating high satisfaction and ease of use. With features for tracking nutritional intake and expert consultations.

Desain dan Implementasi Sistem Kontrol Suhu dan Kelembaban di Greenhouse dengan Pendekatan Fuzzy

2025-01-15T06:25:15+01:00

This research aims to develop an automatic control system for temperature and humidity in a greenhouse using the NodeMCU ESP 8266 microcontroller in the context of Smart Farming. Optimal greenhouse environmental control is crucial for increasing crop productivity. The system collects temperature and humidity data from sensors, with control using Fuzzy logic rules designed in Matlab. Test results show that the NodeMCU functions well at an average voltage of 4.7 V. The soil moisture sensor provides readings corresponding to soil conditions: dry (12%-14%), wet (66%-71%), and after watering (39%-42%). The DHT11 sensor recorded an average temperature of 33°C and air humidity of 55%. This system provides an automatic solution responsive to environmental changes, creating optimal conditions for plant growth, while improving efficiency and productivity in modern agriculture. This research contributes to the development of sustainable smart farming.

Sistem Informasi Kelayakan TORA Berbasis Web di Kabupaten Cianjur

2025-06-19T18:26:11+02:00

Agrarian reform is a strategic program aimed at improving land ownership distribution and enhancing community welfare. One crucial stage in this program is determining the eligibility of Tanah Objek Reforma Agraria (TORA), which requires an objective and structured selection method. This study develops a web-based system that implements the Simple Additive Weighting (SAW) method to assess TORA eligibility more efficiently and transparently. The SAW method is chosen because it provides systematic calculations by considering various established criteria, such as land legal status, land use, and social and economic potential. This system enables policymakers to manage and analyze data more effectively, thereby supporting accurate and fair decision-making in the implementation of agrarian reform. The study results indicate that a web-based approach using the SAW method can improve accuracy and efficiency in determining TORA eligibility, offering a more structured solution for agrarian reform programs in Indonesia.

Evaluasi Sistem Pelacakan Resi Pada Website Shopee Express Menggunakan Pendekatan Pieces

2025-06-25T16:57:16+02:00

This study uses the PIECES framework, which has six important aspects: Performance, Information, Economy, Control, Efficiency, and Service, to assess the shipment tracking system on the Shopee Express website. Direct observation and system testing serve as the foundation for the study's qualitative descriptive methodology. The system's accuracy in displaying real-time tracking information and its performance on stable networks are demonstrated by the results. However, a number of drawbacks are discovered, including restricted service responsiveness on sluggish connections, no progress feedback during data loading, and no input validation. The system lacks user assistance features like history tracking and interactive guidance, despite being cost-effective and available without a login. The PIECES framework is a useful tool for determining web-based logistic tracking systems' advantages and shortcomings.

Analisis Perbandingan Kinerja Algoritma Klasifikasi Data Menggunakan Metode K-NN, Naive Bayes, dan Decision Tree pada Dataset UCI Iris

2025-06-28T17:14:41+02:00

Data classification is one of the important techniques in data mining and machine learning, which is widely used to group data into certain classes. This study aims to analyze and compare the performance of three classification algorithms, namely K-Nearest Neighbor (K-NN), Naive Bayes, and Decision Tree, in classifying Iris data from the UCI Machine Learning Repository. This dataset consists of 150 data with four feature attributes and three target classes. Testing was carried out using the cross-validation method with a k-fold approach of 10 folds. The results of the performance evaluation were measured using the metrics of accuracy, precision, recall, and f1-score. Based on the test results, the K-NN algorithm showed the highest accuracy rate of 96.67%, followed by Decision Tree at 95.33%, and Naive Bayes at 94.00%. These findings indicate that choosing the right classification algorithm can affect the success rate in the data classification process.

The Impact of Aurora Supercomputer in Scientific Area and the Development Analysis

2025-07-03T04:36:47+02:00

Supercomputers are crucial in solving complex scientific and industrial computing due to its tremendous computational power in enabling large- scale simulations, scientific research, and advancement in various scientific fields. This study is conducted on Aurora supercomputer, a powerful exa-scale supercomputer designed for intricate computing tasks, such as climate modeling, intense simulations, and AI and machine learning. By making use of the literature review approach, we analyze the capabilities and impact of Aurora on the scientific environment. Our research suggests that Aurora is capable of significantly enhancing performance on processing data, surpassing supercomputers such as Frontier and Fugaku. Furthermore, we discuss Aurora's impact on driving groundbreaking research across multiple scientific domains and its real-world applications such as drug discovery driven by AI and machine learning. The result highlights that Aurora marked a remarkable milestone in revolutionizing computational research and further research can show the true power of the Aurora supercomputer