| Publisher: Al-Kindi Center for Research and Development (Total: 14 journals) [Sort by number of followers]
| ||||||||||||||||||||||||||||||
Similar Journals
|
Journal of Computer Science and Technology Studies
Number of Followers: 1  Open Access journalISSN (Online) 2709-104X Published by Al-Kindi Center for Research and Development  [14 journals]
|
- Deep Learning-Based COVID-19 Detection from Chest X-ray Images: A
Comparative Study
Authors: Duc Minh Cao; Md Shahedul Amin, Md Tanvir Islam, Sabbir Ahmad, Md Sabbirul Haque, Md Abu Sayed, Md Minhazur Rahman, Tahera Koli
Abstract: The COVID-19 pandemic, caused by the SARS-CoV-2 virus, has rapidly spread across the globe, leading to a significant number of illnesses and fatalities. Effective containment of the virus relies on the timely and accurate identification of infected individuals. While methods like RT-PCR assays are considered the gold standard for COVID-19 diagnosis due to their accuracy, they can be limited in their use due to cost and availability issues, particularly in resource-constrained regions. To address this challenge, our study presents a set of deep learning techniques for predicting COVID-19 detection using chest X-ray images. Chest X-ray imaging has emerged as a valuable and cost-effective diagnostic tool for managing COVID-19 because it is non-invasive and widely accessible. However, interpreting chest X-rays for COVID-19 detection can be complex, as the radiographic features of COVID-19 pneumonia can be subtle and may overlap with those of other respiratory illnesses. In this research, we evaluated the performance of various deep learning models, including VGG16, VGG19, DenseNet121, and Resnet50, to determine their ability to differentiate between cases of coronavirus pneumonia and non-COVID-19 pneumonia. Our dataset comprised 4,649 chest X-ray images, with 1,123 of them depicting COVID-19 cases and 3,526 representing pneumonia cases. We used performance metrics and confusion matrices to assess the models' performance. Our study's results showed that DenseNet121 outperformed the other models, achieving an impressive accuracy rate of 99.44%.
PubDate: Tue, 28 Nov 2023 00:00:00 +000
- Improving the Efficiency of Distributed Utility Item Sets Mining in
Relation to Big Data
Authors: Arkan A. Ghaib; Yahya Eneid Abdulridha Alsalhi, Israa M. Hayder, Hussain A. Younis, Abdullah A. Nahi
Abstract: High utility pattern mining is an analytical approach used to identify sets of items that exceed a specific threshold of utility values. Unlike traditional frequency-based analysis, this method considers user-specific constraints like the number of units and benefits. In recent years, the importance of making informed decisions based on utility patterns has grown significantly. While several utility-based frequent pattern extraction techniques have been proposed, they often face limitations in handling large datasets. To address this challenge, we propose an optimized method called improving the efficiency of Distributed Utility itemsets mining in relation to big data (IDUIM). This technique improves upon the Distributed Utility item sets Mining (DUIM) algorithm by incorporating various refinements. IDUIM effectively mines item sets of big datasets and provides useful insights as the basis for information management and nearly real-time decision-making systems. According to experimental investigation, the method is being compared to IDUIM and other state algorithms like DUIM, PHUI-Miner, and EFIM-Par. The results demonstrate the IDUIM algorithm is more efficient and performs better than different cutting-edge algorithms.
PubDate: Fri, 24 Nov 2023 00:00:00 +000
- Block Diagonalization in the 5G SA Network
Authors: Mohamed Mokrani; Messaoud Bensabti
Abstract: In this paper, we did programming regarding the Block diagonalization technology in the 5G standalone SA network, in this program, we have created a 5G site with 16 antennas(minimum of Massive MIMO) and 4 active users equipped of 4 antennas, this system is called Multi Users Massive MIMO system, the link that was chosen is the downlink,we have calculated the maximum throughput in the 5G downlink where we have obained a value of 1673864 b/ms, this value is divided by the number of Massive MIMO layers which worth 16 to get a transport block size of 104616 b/ms (no Cyclic redundancy check CRC). The Block Error rate BLER is null (no detection of errors in reception) because we are in the case of no crc and no channel coding (uncoded transmission), the signal of each user among 4 to be transmitted consists of 4 vectors, each vector has a length of 52308 that corresponds to the number of symbols which are the outputs of Quadrature Phase Shift Keying QPSK Mapping Operation. The received signal at each user equipment UE has a form which can be represented by the multiplication of preconding matrix of this UE with the channel matrix between this UE and the 5G site plus the noise received at the antennas of this UE. the results show that the product of channel gain between UE and the 5G site(known in emission) with the precoding matrix of the other UE gives a matrix which composes of imaginary elements each of which has a real part and imaginary part which both tend to zero(the inter users interferences IUI is canceled). The results show also that when the Signal to Noise Ratio SNR increases(several transmissions) the Bit Error Rate decreases.
PubDate: Sat, 18 Nov 2023 00:00:00 +000
- Generative AI Model for Artistic Style Transfer Using Convolutional Neural
Networks
Authors: Jonayet Miah; Duc Minh Cao, Md Abu Sayed, Md Sabbirul Haque
Abstract: Artistic style transfer, a captivating application of generative artificial intelligence, involves fusing the content of one image with the artistic style of another to create unique visual compositions. This paper presents a comprehensive overview of a novel technique for style transfer using Convolutional Neural Networks (CNNs). By leveraging deep image representations learned by CNNs, we demonstrate how to separate and manipulate image content and style, enabling the synthesis of high-quality images that combine content and style in a harmonious manner. We describe the methodology, including content and style representations, loss computation, and optimization, and showcase experimental results highlighting the effectiveness and versatility of the approach across different styles and content.
PubDate: Fri, 17 Nov 2023 00:00:00 +000
- Empirical Study on the Relationship between Users’ Mental Model and
Purchase Intention of VIP Subscription: Evidence from Image Processing App
in China
Authors: Yuguo Gao
Abstract: With the Internet entering the inventory stage, subscription services have become a major trend in the industry. As a technology company driven by artificial intelligence and with beauty as core, Meitu has launched VIP subscription services in several image processing applications. By December 2022, the number of VIP members grew to about 5.6 million, becoming a new engine for the company to open up more business space. At present, there is few research in academia on the VIP subscription intention of image processing APP. Combining the characteristics and usage experience of image processing APP, this thesis constructed the research model by introducing the concept of user’s mental model in the technology acceptance model. Using the structural equation modeling method, the hypothetical model and the relationship between critical variables was validated. With SPSS28.0 and AMOS24.0 software, the confirmatory factor analysis, exploratory factor analysis and structural equation modeling was conducted. The results indicate that both quality of system interface and quality of subscription service positively influence user’s mental model; mind model of users influences purchase intention through the direct path. At the same time, it also influences purchase intention through perceived ease of use and perceived usefulness, and the chain mediating path between them. Based on the findings, this thesis claims that Meitu should increase the investment in scientific research; it should not only focus on the optimization of system interface design, pay attention to the professionalism and personalized upgrade of subscription services, but also dig deeper into users’ needs and occupy their minds. At the same time, Meitu App should promote the subscription model with precise positioning and tiered payment, so as to increase users’ intention of subscription.
PubDate: Fri, 17 Nov 2023 00:00:00 +000
- Enhancing Traffic Density Detection and Synthesis through Topological
Attributes and Generative Methods
Authors: Jonayet Miah; Md Sabbirul Haque, Duc Minh Cao, Md Abu Sayed
Abstract: This study investigates the utilization of Graph Neural Networks (GNNs) within the realm of traffic forecasting, a critical aspect of intelligent transportation systems. The accuracy of traffic predictions is pivotal for various applications, including trip planning, road traffic control, and vehicle routing. The research comprehensively explores three notable GNN architectures—Graph Convolutional Networks (GCNs), GraphSAGE (Graph Sample and Aggregation), and Gated Graph Neural Networks (GGNNs)—specifically in the context of traffic prediction. Each architecture's methodology is meticulously examined, encompassing layer configurations, activation functions, and hyperparameters. With the primary aim of minimizing prediction errors, the study identifies GGNNs as the most effective choice among the three models. The outcomes, presented in terms of Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE), reveal intriguing insights. While GCNs exhibit an RMSE of 9.25 and an MAE of 8.2, GraphSAGE demonstrates improved performance with an RMSE of 8.5 and an MAE of 7.6. Gated Graph Neural Networks (GGNNs) emerge as the leading model, showcasing the lowest RMSE of 9.2 and an impressive MAE of 7.0. However, the study acknowledges the dynamic nature of these results, emphasizing their dependency on factors such as the dataset, graph structure, feature engineering, and hyperparameter tuning.
PubDate: Thu, 16 Nov 2023 00:00:00 +000
- A Study of Organizational Changes that Occur to the Adoption of Cloud
Computing Technologies in Organizations: Ministry of Communication and
Information Technology in Afghanistan
Authors: Mohammadullah Shirpoor; Nasrullah Ranimi, Asmatullah rashidi
Abstract: Cloud computing services such as file storage and big data analysis provide cost effective, secure, flexible and reliable services to their users; however, their advantages, the adoption of many cloud services is still limited, and many organizations are unsure of adopting cloud technologies for various reasons this study using a systematic review of the factors influencing organizational regarding the adoption of cloud computing technologies, categorize and compare these factors and show that much of the literature has highlight the technical aspects of technology adoption, such as cloud security further show that factors such as top management support, relative advantage, cloud complexity, and competitive pressure are the most important factors affecting organizational attitudes toward cloud technology adoption. Furthermore, analysis of interview data collection techniques showed that cloud computing technologies affect the structure, size, tasks and work processes of organizations. These variables change at different levels. The findings showed that IT jobs have the greatest impact on cloud computing readiness and performance. Additionally, the results showed that organizations that adopt cloud technologies integrated some departments, increased work speed, removed some duplicated steps, overcame management changes, centralized IT works and removed some traditional hierarchical parts.
PubDate: Wed, 08 Nov 2023 00:00:00 +000
- IoT-based Electrical Power Recording using ESP32 and PZEM-004T
Microcontrollers
Authors: Kadek Amerta Yasa; I Made Purbhawa, I Made Sumerta Yasa, I Wayan Teresna, Aryo Nugroho, Slamet Winardi
Abstract: The electricity usage recording system in Indonesia still uses conventional kWh meters. Electricity usage is recorded by officers who visit customers' homes every month. This results in the electricity company having to provide employees who become a burden on the company's costs. Technological advances enable convergence between communication channels and various things. A technology known as the Internet of Things (IoT) allows customer kWh meters to be recorded in real-time. This research aims to create an Internet of Things (IoT)-based kWh meter that can make it easier for electricity companies to monitor each customer's electricity usage. The IoT kWh meter created can be monitored and controlled from a remote location in real-time. If there is a change in load usage, it will be monitored directly via a mobile device because the kWh meter is directly connected to the internet network and cloud server. To determine the functionality of the tool being made, several tests were carried out, such as a) sensor testing, b) LED indicator, buzzer, and relay testing, c) OLED display testing, d) Firebase database testing, and e) load testing. The test results obtained are used to calculate the error of the tool made with a comparator, and the results show that the percentage of voltage error with different loads is very small, namely 0.35% and 1.45%. This research produced a prototype using ESP32 and PZEM-004T, which is so accurate that it is recommended for recording electrical power, which can reduce the burden on operational costs for electricity companies.
PubDate: Wed, 08 Nov 2023 00:00:00 +000
- Application of Data Mining with K-Nearest Neighbors Algorithm for Shallot
Price Prediction
Authors: Yuana Inka Dewi Br Sinulingga; Donny Avianto
Abstract: Shallots are an important and widely consumed bulb crop in Indonesia, both for medicinal and culinary purposes. However, shallot yield is substantially affected by its supply, often leading to significant price fluctuations that greatly impact consumers and producers, especially farmers. Farmers who cannot accurately predict shallot prices often incur losses when selling to shallot distributors. If this problem is not resolved, it may discourage farmers from cultivating shallots. Therefore, a prediction system is needed to forecast shallot prices in the future, thus helping farmers make the right decisions. This research uses the K-Nearest Neighbors (KNN) algorithm for shallot price prediction. KNN classifies data into specific categories based on the closest distance to a set of k patterns for each category, using the Euclidean distance formula to calculate the distance. The dataset consists of 303 entries with five features: farmer price, seller price, retail price, seed price, and yield. The test results of the Shallot Price Prediction System in North Sumatra Province, Indonesia, using the K-Nearest Neighbors Algorithm, showed the best performance when using 80% training data and 20% testing data, with a value of k=2, resulting in a Mean Absolute Error (MAE) of 25,786 and a Mean Squared Error (MSE) of 72. This system empowers farmers to predict the future price of shallots before selling their crops to distributors.
PubDate: Wed, 01 Nov 2023 00:00:00 +000
- Improved Neural Network-Based System for Early and Accurate Diagnosis of
Alzheimer Disease
Authors: Spogmay Yousafzai; Gul Zaman khan, Sajad Ulhaq, Areebah, Muhammad Rabbi Butt
Abstract: Alzheimer's disorder is a neurological condition that develops over time and mainly impacts cognitive processes like memory, thought, and behavior. It is one of the most typical reasons for dementia, a syndrome marked by a loss of cognitive ability that interferes with individual daily activities. Recent techniques for diagnosing Alzheimer's illness frequently combine positron emission tomography (PET) scans with magnetic resonance imaging (MRI), which can identify mutations in the brain caused by the illness, such as the buildup of beta-amyloid plaques and tau tangles. Furthermore, analysis of blood samples and cerebrospinal fluid is also a widely used method for the diagnosis of Alzheimer’s disease. Machine learning and deep learning-based techniques play a vital role in examining complex structures in brain images and other data, contributing to the timely and precise identification of Alzheimer's disease. Artificial intelligence-based techniques can help prompt detection and treatment, leading to more efficient care for Alzheimer's disease. This study uses convolutional neural networks (CNN) with MRI-based datasets for early and accurate diagnosis of Alzheimer’s disease. The proposed approach has shown excellent results in AD diagnosis.
PubDate: Thu, 26 Oct 2023 00:00:00 +000
- Analyzing Supporting and Inhibiting Factors in the Optimization of
E-Government in Pontianak City
Authors: Bagus Pramono Rusadi; Andi Fitri
Abstract: This study investigates the potentialities and challenges of e-government optimization in Pontianak City, Indonesia, amidst the pressing demands for efficient and quality public services fueled by globalization. Despite the increased adoption of technology and the high penetration of internet and mobile devices in Pontianak, the implementation of e-government remains suboptimal, contributing to weak governance and limited public services. Employing a qualitative research method with a descriptive approach, this study systematically explores the tangible, intangible, and highly intangible challenges inhibiting e-government optimization, such as inadequate IT infrastructure, financial constraints, limited human resource capabilities, and a lack of standardization and integration in content development. However, the presence of regulations, implementing institutions, and advancements in developer competence in content development emerge as supporting elements for e-government realization. Furthermore, the study identifies connectivity issues, low technological literacy, and insufficient budgets as critical roadblocks. The findings underscore the necessity for multifaceted and comprehensive strategies to overcome the identified barriers and unlock the full potential of e-government in enhancing governance and public service delivery in Pontianak and similar settings, thereby contributing to the literature on e-government and offering valuable insights for stakeholders and policy-makers aiming to foster digital era governance.
PubDate: Sat, 14 Oct 2023 00:00:00 +000
- Financial Analysis Dashboard Application for Stock Exchange Listed
Companies
Authors: Florina Covaci; Dragoș Boscan
Abstract: The current paper aims to outline the development of a web application to streamline the process of analyzing listed companies in a simpler, more concise and more user-friendly way, helping financial analysts make better decisions when placing a trade. The application offers users the ability to obtain financial analysis through a single company search as well as the ability to record transactions to account for price changes. The fundamental analyzes that the application offers start with the analysis of solvency, cash conversion cycle, performance, positioning, liquidity and bankruptcy risk, using public financial data of the companies as well as the current situation of the news and changes they have had place on the market in the last 24 hours. In the case of technical analysis, we can identify indicators that follow the analysis of the share price, the movements, the trend as well as the trading volume.
PubDate: Fri, 13 Oct 2023 00:00:00 +000
- Detection of Product Cost for Blind People Based on Android Application
Authors: Rakibuzzaman; Pantho Datta, Tonny Roy, Nabil Hayat, MD Fazley Rabbe
Abstract: Blindness refers to a condition in which a person suffers from disturbance or interruption in the sense of vision. Blind individuals are generally classified into two major groups: complete blindness and visual impairment. There have been numerous initiatives and aids developed to support individuals with disabilities, including those with vision impairments. One such aid that could be implemented is an application designed to make shopping easier for the blind. It is widely acknowledged that individuals with disabilities often require assistance from others, and the development of assistive technologies presents an opportunity for blind individuals to become more self-sufficient. The primary objective of this research is to develop an Android application called “See in Me” that can aid blind individuals in shopping more easily, particularly in Bangladesh, where such facilities are currently unavailable. The “See in Me” application will listen to the user's voice instructions, after which the camera within the app will be activated. The user is able to scan any packaged product, which will be recognized by the Android application. The app will provide the user with information such as the product's name and price via voice mode. If the user wants to buy the scanned product, they can add it to the cart by shaking the phone or using the “OK” voice instruction. This process allows users to add more products to their cart. After completing the purchase, the user can use the “OK” instruction to see the total price and list of products added to the cart and can invoke the “Delete” instruction to remove the product from the cart. To ensure that the “See in Me” application is effective, it is crucial to develop accurate and reliable voice recognition technology that can interpret the user's voice instructions without errors. The application should also have a user-friendly interface with clear instructions and feedback to guide the user through the shopping process. It is also important to ensure that the application is compatible with different devices and operating systems, making it accessible to a larger number of users. Testing the application with a diverse group of users, including individuals with varying degrees of visual impairment, can help to identify any usability issues and ensure that the application meets the needs of its target audience.
PubDate: Mon, 02 Oct 2023 00:00:00 +000
