Journal of Computer Networks, Architecture and High Performance Computing
Not a member yet
473 research outputs found
Sort by
Enhancing Multi-Layer Perceptron Performance with K-Means Clustering
Get PDFMachine learning plays a crucial role in identifying patterns within data, with classification being a prominent application. This study investigates the use of Multilayer Perceptron (MLP) classification models and explores preprocessing techniques, particularly K-Means clustering, to enhance model performance. Overfitting, a common challenge in MLP models, is addressed through the application of K-Means clustering to streamline data preparation and improve classification accuracy. The study begins with an overview of overfitting in MLP models, highlighting the significance of mitigating this issue. Various techniques for addressing overfitting are reviewed, including regularization, dropout, early stopping, data augmentation, and ensemble methods. Additionally, the complementary role of K-Means clustering in enhancing model performance is emphasized. Preprocessing using K-Means clustering aims to reduce data complexity and prevent overfitting in MLP models. Three datasets - Iris, Wine, and Breast Cancer Wisconsin - are employed to evaluate the performance of K-Means as a preprocessing technique. Results from cross-validation demonstrate significant improvements in accuracy, precision, recall, and F1 scores when employing K-Means clustering compared to models without preprocessing. The findings highlight the efficacy of K-Means clustering in enhancing the discriminative power of MLP classification models by organizing data into clusters based on similarity. These results have practical implications, underlining the importance of appropriate preprocessing techniques in improving classification performance. Future research could explore additional preprocessing methods and their impact on classification accuracy across diverse datasets, advancing the field of machine learning and its application
Implementation of Naïve Bayes Method Diagnosing Diseases Nile Tilapia
Get PDFThe Nile tilapia, also known as Oreochromis niloticus, was a freshwater fish species first produced in East Africa in 1969. It became a popular aquaculture fish in freshwater ponds across Indonesia. Besides its delicious taste, the Nile tilapia is rich in nutrients essential for human health. However, cultivating Nile tilapia was challenging due to frequent bacterial diseases. These diseases often led to mass fish deaths, causing financial losses, especially for new fish farmers. The rapid spread of diseases emphasized the need for prompt intervention to prevent further losses. Farmers needed adequate knowledge about Nile tilapia diseases, but often struggled to absorb information provided by the government. Hence, the presence of experts or veterinarians was crucial in assisting farmers to address these issues. Farmers of Nile tilapia sought assistance from experts or veterinarians, but this was not easy. It involved substantial costs and time, while quick intervention was necessary to mitigate losses. The solution proposed was the development of an expert system for diagnosing and treating Nile tilapia diseases. Thus, an expert system was built to assist fish farmers in identifying fish diseases and their treatments by implementing the naïve Bayes method. The expert system transferred human knowledge to computers, enabling them to solve problems like experts, thereby making expert knowledge accessible to non-experts. Naïve Bayes was implemented to determine the highest probability based on input symptoms. This research used five test data samples to apply the naïve Bayes method to diagnose Nile tilapia diseases, resulting in an accuracy rate of 80%. Therefore, the implementation of naïve Bayes in diagnosing Nile tilapia diseases is considered reasonably effective
Analysis of the Multi Objective Optimization by Ratio Analysis (MOORA) Method in Determining Pilot Areas at PT. XYZ
No full textThis research analyzes the application of the Multi Objective Optimization by Ratio Analysis (MOORA) method model in determining the Pilot Area at PT XYZ. This method is used to evaluate various performance criteria, including customer satisfaction, productivity, service quality, and operational efficiency. Currently, the Pilot Area assessment and selection process at PT XYZ is still done manually, which causes a lack of accuracy and efficiency. MOORA was chosen for its ability to handle multi-criteria decision-making problems more systematically and objectively. The analysis results showed that Alternative Area 7 obtained the highest final score of 0.39, placing it as an area with superior performance. The application of MOORA is proven to improve accuracy and efficiency in the Pilot Area determination process, providing a more objective basis for decision-making. By using MOORA, PT XYZ can evaluate area performance more comprehensively and accountably. This research recommends that PT XYZ implement the MOORA method thoroughly and conduct periodic evaluations of the methods used. For theory development, PT XYZ can add specific evaluation criteria according to company needs. The implementation of these suggestions is expected to improve the quality of service and competitiveness of PT XYZ in the global market. Further research is expected to compare MOORA with other methods to strengthen the validity of the results. Thus, this research not only provides a practical contribution to PT XYZ but also adds academic insight into the application of multi-criteria optimization methods in the context of performance management and service improvement
Web-Based Archiving Information System At The Ministry of Public Works and Housing Province of Jambi
Get PDFThis research was conducted at the Jambi Province Public Works and Public Housing Office (PUPR). This research is entitled "Design of a Website-Based Archiving Information System at the Jambi Province Public Works and Public Housing Office (PUPR)". The purpose of this research is to document the process of designing an information system that functions to provide information to the PUPR Office of Jambi Province. This research was conducted qualitatively by conducting interviews and observations with the PUPR Office of Jambi Province to complete the research. The document filing system of the Jambi Province PUPR Office was designed using the prototype method. As a result of the analysis and observations made, it can be concluded that one effective way to facilitate the work of employees in finding documents is to build a web-based archiving information system. This system is expected to produce information quickly, precisely, and accurately without reducing the value of the information itself
Optimizing Decision-Making for Aid Allocation in Underdeveloped Regions Using the MOORA Method
Get PDFThe allocation of assistance for the Family Hope Program is a process that requires precision to ensure that assistance is given to those most in need. This research develops a Decision Support System (DSS) using the Multi-Objective Optimization on the Basis of Ratio Analysis (MOORA) method for optimizing the selection of beneficiaries in disadvantaged villages which includes criteria used including education, toddlers, pregnant women, disabilities, elderly, income, employment, number of dependents, and house size. Each criterion is normalized and given a weight according to its level of importance. The results show that alternative A2 has the highest optimization value with Yi of 0.254, followed by A8 (0.208) and A5 (0.204). In contrast, alternatives A3 (0.029) and A10 (0.035) have the lowest optimization value. Matrix normalization and criteria weights show the significant influence of the criteria of education, pregnant women, elderly, income, number of dependents, and house size in the selection process. The implementation of DSS with the MOORA method is proven to increase efficiency and accuracy in the selection process of Family Hope Program beneficiaries, reduce subjective errors, and ensure assistance is channeled to those who really need it. Therefore, the MOORA method is recommended as an effective tool to optimize social assistance allocation, increase transparency, and reduce bias in decision-making
Optimisation of Inventory Management Through Time Series Analysis of Inventory Data with Double Exponential Smoothing Method
Get PDFStock forecasting is very useful for companies in knowing the trend of inventory needed in the next period, with time series data often forecasting can be a solution in supporting decision making. Excess or lack of stock of goods is often caused by a less than optimal record management process and often relies on personal intuition. In this study, the Double Exponential Smoothing method is applied in analyzing time series data and forecasting stock data. This method is used because it is in accordance with the company's sales data which is up and down. In addition, this forecasting calculation does not escape the error rate of forecasting calculations, therefore this system is also supported by the MAD (Mean Absolute Deviation), MSE (Mean Square Error) and MAPE (Mean Absolute Percentage) methods to calculate the error rate of the forecasting results. The forecasting results show that this method is able to provide fairly accurate predictions with a MAD value of 5.2475, MSE of 43.009, and MAPE of 26.307%. By using DES, companies can perform better stock planning, reduce the risk of over- or under-stocking, and improve inventory management efficiency. The DES method is proven to be flexible and easy to implement in computerized information systems, so it is recommended to be used more widely in corporate inventory management
Reverse Engineering for Static Analysis of Android Malware in Instant Messaging Apps
Get PDFMalware poses a significant threat to Android devices due to their high prevalence and vulnerability to attacks. Analyzing malware on these devices is crucial given the persistent and sophisticated threats targeting Android users. Static analysis of Android malware is a key approach used to detect malicious software without executing the application. This method involves meticulously examining the application's source code or binaries to identify signs of suspicious or harmful activities. The research methodology consists of three stages. The first stage involves collecting malware samples spread through instant messaging applications. The second stage employs reverse engineering, where APK files are decompiled to extract their contents. Following this, a static analysis is conducted, focusing on the AndroidManifest.xml file and the source code to identify the behavior and potential threats posed by the malware. The static analysis results revealed that Android malware often requests sensitive permissions to access personal data, such as receiving, reading, and sending SMS, as well as accessing location and contacts. Further analysis uncovered that after acquiring this data, the malware transmits it to the Telegram API via authenticated HTTP requests using specific tokens and chat_ids. These findings highlight that the permissions requested by the malware are designed to clandestinely collect and export personal data, posing a severe threat to the privacy and security of Android users
Improving Information Security with Machine Learning
Get PDFThe study Improving Information Security with Machine Learning explores the fusion of machine learning methodologies within information security, aiming to fortify conventional protocols against evolving cyber threats. By conducting a comprehensive literature review and empirical analysis, this scholarly endeavor highlights the efficacy of machine learning in anomaly detection, threat identification, and predictive analytics within security frameworks. Through practical demonstrations, such as z-score-based anomaly detection in network traffic data and NLP-based email security systems, the study illustrates the practical applications of machine learning techniques. Additionally, it delves into the mathematical underpinnings of predictive analytics and the architecture of neural networks for malware detection. However, while showcasing the transformative potential of machine learning, the study also confronts significant challenges. Ethical, legal, and privacy considerations emerge prominently, emphasizing the need for regulations addressing algorithmic biases, ethical dilemmas, and data protection. Moreover, the study emphasizes the practical challenges of scalability, interpretability, continual adaptation to evolving threats, and the harmonious interaction between human expertise and machine intelligence. By offering practical recommendations and future research directions, this scholarly exploration aims to empower researchers, practitioners, and policymakers in navigating the complex intersection of machine learning and information security, thereby fostering innovation and comprehension in this evolving domain
E-Commerce Application with Web Engineering Method Website Based
No full textCV Jon Indo merupakan perusahaan yang bergerak di bidang material pabrik sebagai supplier, kontraktor, dan stockist yang memperdagangkan material pabrik lainnya. Untuk bersaing secara global, salah satu alat atau tools yang dapat menjangkau pasar yang diharapkan dalam hal ini adalah peningkatan penjualan produk. CV. Jon Indo belum memiliki sistem pemesanan yang mampu menjangkau banyak konsumen yaitu masih menjual barangnya melalui WhatsApp dan merekapitulasi data penjualan secara manual sehingga terdapat permasalahan seperti kesalahan penjadwalan pengiriman barang, serta kesalahan input produk. pesanan sehingga menyebabkan penjualan menurun. Penelitian ini bertujuan untuk merancang sebuah aplikasi yang ditampilkan pada CV Jon Indo untuk meningkatkan volume penjualan dan memperluas pemasaran produk CV Jon Indo. Sistem yang dibangun mempunyai fitur promosi dan penambahan CTA (call to action) pada website, kemudian pada sistem pengiriman barang peneliti membuat fitur tracking dan fitur permintaan barang sesuai keinginan pelanggan. Aplikasi berbasis web ini dirancang dengan menggunakan metode Research and Development (RnD) dan Metode Web Engginering sebagai perancangan aplikasi E-Commarece. Bahasa pemrograman yang digunakan untuk implementasi E-comarece menggunakan bahasa pemrograman PHP dan database MySql. Aplikasi yang dibuat dapat digunakan untuk memudahkan penjualan, pembelian dan penyebaran informasi produk terbaru
Utilization of Data Analytics to Enhance Operational Efficiency in Manufacturing Companies
No full textIn the digital era, manufacturing industries confront challenges like heightened global competition and intricate production processes, urging them to boost efficiency and productivity. Amidst these circumstances, Big Data emerges as a pivotal opportunity to enhance manufacturing performance. Big Data, characterized by vast volumes of data, utilizes advanced data mining to machine learning techniques for analysis. Data analytics, an interdisciplinary field, profoundly impacts manufacturing operations, enabling deeper insights into production processes. By analyzing production data, companies identify inefficiencies, streamline workflows, and enhance operational efficiency and productivity. Predictive maintenance through sensor data analysis prevents machine failures, while logistics data analysis optimizes supply chains and inventory management, reducing costs and enhancing competitiveness. However, implementing Big Data analytics presents challenges such as rapid data growth, diverse data sources, real-time insights, skill shortages, and data fragmentation. Overcoming these hurdles requires robust technology, skilled personnel, and effective data management strategies. Examples of Big Data analytics applications include customer behavior analysis by Amazon and Netflix, fraud detection in insurance, and urban mobility optimization. Success factors in data analytics implementation include effective data-driven communication, technology integration, and skill enhancement. In conclusion, implementing Big Data Analytics in manufacturing promises significant benefits in operational efficiency, product quality, and competitiveness. Overcoming challenges necessitates robust strategies and consideration of ethical and security issues, ensuring responsible data usage. With a deep understanding of Big Data Analytics, manufacturing companies can leverage this technology to achieve higher efficiency and competitiveness in the global market