International Journal of Communication Networks and Information Security (IJCNIS)
Not a member yet
1021 research outputs found
Sort by
Optimizing Wireless Sensor Network Localization: Hybrid Natural Inspired with Bat Swarm Algorithm
No full textIn wireless sensor networks, the data received from sensor nodes is processed and communicated to the next node or cluster head. Sensed data are meaningful only when location information is embedded with the data. Therefore, the sensor should be able to estimate its location information and embed the location information with the sensed data. This research work proposes three hybrid nature-inspired localization optimization algorithms for location estimation of sensor nodes, namely the Hybrid genetic –Bat Localization Algorithm (GA-BAT LA), Hybrid Bat- Particle Swarm Optimization Localization Algorithm (BAT-PSO LA), and Hybrid genetic -Bat- Particle Swarm Optimization Localization Algorithm (GA-BAT-PSO LA). The localization algorithms are developed, implemented, and compared to ensure the accuracy of the estimation of location information. A fixed number of location-known anchor nodes with a variable location-unknown sensor node are considered. The algorithms are compared for a minimum number of iterations, average number of iterations, standard deviation, maximum number of iterations, time complexity, and accuracy to estimate the location of the sensor nodes
Optimizing Deep Residual Networks: Incorporating Separable Convolutions into ResNet50 Architecture
No full textCorn crop diseases have a significant impact on agricultural productivity, making efficient and accurate detection critical for timely intervention. This paper introduces a novel approach for classifying corn crop diseases using a hybrid deep learning model that combines ResNet50 with separable convolutions. Separable convolutions decompose standard convolutions into two smaller operations—depthwise and pointwise convolutions—thus reducing computational complexity while preserving high accuracy by processing spatial and channel-wise information separately. The proposed hybrid model is trained to categorize images into four classes: healthy, Cercospora leaf spot, common rust, and Northern leaf blight, using a dataset of 38,520 images. The model's performance is compared with transfer learning approaches like Xception, ResNet50, MobileNetV2, and EfficientNetB0. Experimental results demonstrate that the proposed model achieves an accuracy of 98.3%, while significantly reducing computational complexity by decreasing the number of parameters, approximately 61.3% less than those in ResNet50
Money Management and Artificial Intelligence (Ai): The Implications
No full textThe rapid convergence of artificial intelligence (AI) and money management is examined in this study, with particular attention paid to the significant effects this integration has on the finance industry. AI is changing traditional financial practices by providing new tools for asset management, trend prediction, and improved decision-making. These tools are becoming more complex as AI technology advance. In order to enhance risk assessment, automate trading, and optimize portfolio management, this research looks at the applications of AI-driven algorithms and machine learning models. The study also highlights the transition from human-led guidance to AI-powered solutions that provide individualized, data-driven insights in the discussion of the effects of AI on financial advising services. The study also discusses the possible hazards and ethical issues that come with using AI to money management, such as worries about algorithmic bias, data privacy, and the transparency of judgments made by AI. This article attempts to give a thorough knowledge of how artificial intelligence (AI) is changing the money management environment and the potential and difficulties it brings for financial institutions and individual investors going forward. It does this by studying current trends and case studies. According to the findings, artificial intelligence (AI) has a lot to offer in terms of efficiency and precision, but it also needs strict control and oversight to guarantee that it is used fairly and ethically
Enhancing Braille Education: Usability, Perception, and Design Considerations of a Graphical User Interface
No full textReceived: 17 Apr 2024
Accepted: 26 Aug 2024
The study examined the usability, perception, and design considerations of a Graphical User Interface (GUI) aimed at enhancing Braille education. A total of 110 participants, with a mean age of 28.4 years, participated, with a balanced gender distribution and varying education levels. Results from participant perceptions revealed high ratings for the GUI's usability, with mean scores ranging from 4.6 to 4.8 out of 5. Comparisons between blind individuals and educators indicated overall positive perceptions of the GUI, with blind individuals showing slightly higher ratings. Participant feedback highlighted key design features such as clear layout, interactive elements, customizable settings, and real-time feedback. In-depth interviews identified themes emphasizing accessibility and inclusivity, personalized learning experiences, collaboration, feedback mechanisms, and integration with existing curriculum as crucial for effective Braille education using the GUI. These findings underscore the importance of designing user-friendly interfaces that cater to diverse needs, promote engagement, and seamlessly integrate with educational practices to enhance Braille literacy and accessibility for individuals with visual impairments. Such insights can inform the development of future educational technologies, fostering inclusive learning environments and improving educational outcomes for visually impaired individuals
Supervised learning Techniques for Training and Prediction in the Application of Naive Bayes for Social Media Insights
No full textMost existing clustering methods do not fulfill the extraordinary prerequisites of the text document clustering, for example, dealing with high measurement and context-sensitive languages, and providing overlapped clusters. In our day to day lives, Social networking sites such as Instagram, twitter and Facebook play a very major role in connecting people who would otherwise may not be able to stay connected. There is a need of efficient machine learning and information retrieval algorithms to access the required documents from a large set of text documents.
In this paper, a method has been developed to analyse the privacy breach that happens on social media platforms with the help of various machine learning tools such as Naive bayes classification and clustering techniques. We experimented existing methodologies, proposed method and successfully done the analysis work on privacy and data protection for social media data. We have successfully retrieved historic data between 2008-2020 from the Guardian API and performed analysis with the help of different clustering techniques such as K-means clustering and Agglomerative hierarchical clustering algorithms
Protecting Privacy When Using Artificial Intelligence in Retail Systems: Legal Regulations in the Us, Eu And Vietnam
No full textThe rapid development of technical technologies used by offline and online retailers has made consumers' concerns regarding privacy increasingly heightened. With that comes countless tensions for retailers and consumers, trade-offs and compromises just to personalize the shopping experience. In this article, we study privacy regulations when businesses use artificial intelligence in retail systems in the United States - a leading country in artificial intelligence, the European Union - the European Union. The region has the strictest privacy legal policies and Vietnam - a developing country with great potential in artificial intelligence. Based on the mentioned legal regulations, we provide an overview of the data privacy risks that artificial intelligence can bring in the retail industry and the reasonable ways in which Governments of developing countries should react so as not to unintentionally inhibit the development of the digital economy by their strict legal policies. Along with that, our research direction also lays a promising foundation for academic research on privacy rights and a model for building privacy protection laws that take into account the balance between the law and the law. and economic development orientation
Deepincepnet: Disease Detection in Corn or Maize Plant Leaves Using Specim IQ Hyperspectral Imaging and Proposed Dnn Classifiers with Inception Networks
No full textThis research proposes DeepIncepNet, a novel method that combines deep neural networks (DNNs) and hyperspectral imaging to identify illnesses in the leaves of corn or maize plants. The Specim IQ system was utilized to gather hyperspectral imaging data, which encompasses a broad range of wavelengths in spectral information. Using a unique DNN architecture, DeepIncepNet uses Inception Networks (InceptionV3) to classify healthy and damaged maize leaves. To assess the performance of the suggested model, it is compared to well-known architectures as InceptionV3, ResNet-50, and ResNet-101. The results of the experiments suggest that DeepIncepNet achieves greater robustness and accuracy in disease identification, highlighting its potential for early detection and treatment of diseases affecting the maize or corn plant
Growing User Base and Revenue through Data Workflow Features: A Case Study
No full text This research paper presents a comprehensive case study examining the impact of implementing advanced data workflow features on user acquisition, retention, and revenue growth in a Software as a Service (SaaS) company. The study focuses on DataFlow Technologies, a mid-sized data analytics platform provider, and analyzes the effects of three key data workflow improvements implemented over a two-year period. Through a mixed-methods approach combining quantitative analysis of user metrics and qualitative assessment of user feedback, the study demonstrates significant positive correlations between enhanced data workflow capabilities and business growth. The findings provide valuable insights for SaaS companies seeking to leverage data workflow features as a strategic tool for expanding their user base and increasing revenue. The research highlights the importance of user-centric design, scalability, and continuous innovation in developing effective data workflow solutions
A Comparative Study of GPT3.5, GPT4, and Bard-Gemini- Gemini in Addressing Object-Oriented Programming Tasks
No full textLarge Languages Model (LLMs) having shown great potentiality as assistive tools for pupils engaged in programming assignments. Nevertheless, the intricate nature of object-oriented programming (OOP), with its multifaceted requirements for entity identification, relationship establishment, and responsibility delineation, remains a challenge for these models. While LLMs have shown proficiency in introductory programming tasks, their performance in the context of OOP has been under-researched. Current studies targeted towards addressing these gaps by evaluating the efficacy of three prominent LLMs—GPT3.5, GPT4, and Bard-Gemini—in solving current world’s OOPs works utilized inside education-based setting. Answers generated by these models were subsequently validated using an Automatic Assessment Tool (AAT). Results indicate that while the models were frequently capable of producing functional solutions, they often failed to adhere to established OOP best practices. Among the evaluated models, GPT4 exhibited the highest proficiency, followed by GPT3.5, with Bard-Gemini-Gemini demonstrating the least proficiency. Based on these findings, the researcher advocates for a heightened emphasis on code quality when utilizing LLMs for educational purposes and suggests integrating Large Languages Model with AATs like a potential channels towards pedagogical advancement. The study concludes by highlighting the promising capabilities of GPT4 while emphasizing the necessity for continued supervision in the deployment of LLMs within OOP education
Ensemble-based Machine Learning Approach for Automated Software Defect Prediction
No full textIn the tech industry, ensuring software reliability is a critical concern for professionals, often addressed through traditional techniques that rely on prior experience or identifying faulty modules within an application. These methods can be time-consuming and may not always pre-emptively address issues. Automated software defect prediction models, driven by ensemble learning techniques, offer a proactive approach to significantly enhance a software's ability to predict and mitigate defects, leading to more efficient operation, reduced errors, and lower costs. This paper proposes a software defect prediction model based on ensemble learning methods, aimed at maintaining software functionality more effectively. Using established evaluation benchmarks including ten-fold cross-validation, precision, recall, specificity, F1 measure, and accuracy our study evaluates the performance of various machine learning algorithms: Ensemble Learning (EL), Decision Trees (DT), Naive Bayes (NB), Artificial Neural Networks (ANN), and Support Vector Machines (SVM). The results reveal that EL consistently outperforms other models with classification accuracy ranging from 98% to 100%, demonstrating its robustness and superior ability to balance precision and recall across diverse datasets (JM1, CM1, and PC1). Following EL, DT also performs strongly but with slightly lower accuracy, particularly in contexts where interpretability is crucial. NB and ANN show decent results but require careful tuning to achieve optimal performance, while SVM ranks lowest in this analysis. These findings underscore the importance of selecting and implementing appropriate algorithms based on the specific demands of software defect prediction tasks, with EL emerging as the most reliable and robust choice for enhancing software reliability