Fusion: Practice and Applications

The world has become more like a small community thanks to the internet, which connects millions of people, businesses, and pieces of technology for a variety of uses. Because of the significant influence these networks have on our lives, maintaining their efficiency is important, which necessitates addressing issues like congestion. In this study, PI-controller gains are adjusted using a variety of optimization strategies to regulate the nonlinear TCP/AQM model. This controller commits controlled pressured signaling characteristics and modifies computer network congestion. First manual tune PI-Controller are used; then several optimization techniques were used to tune PI-controller gains (Particle Swarm Optimization (PSO), Ant-Colony Optimization (ACO) and Simulated Annealing algorithm (SA)) and then Linear Quadratic Regulator theory are used.  To test the reliability and effectiveness of each of the suggested controllers, several tests utilizing varied network parameter values, different queue sizes, and extra disturbances were conducted. MATLAB was used for all experiments., the results show the superiority of the LQR controller over PI controller with both manual and optimal tuning techniques.

There has yet to be a comprehensive investigation on enhancing the diagnostic accuracy of oral disease using handheld smartphone photographic photos. To overcome the difficulties associated with the automatic detection of oral illnesses, we describe an approach based on smartphone image diagnosis powered by a deep learning algorithm. The centered rule method of image capture was offered as a quick and easy way to get high-quality pictures of the mouth. A resampling method was proposed to mitigate the influence of image variability from handheld smartphone cameras, and a medium-sized oral dataset with five types of disorders was developed based on this approach. Finally, we introduce a recently developed deep-learning network to assess oral cancer diagnosis. On 455 test images, the proposed technique showed an impressive 83.0% sensitivity, 96.6% specificity, 84.3% accuracy, and 83.6% F1. The proposed "center positioning" method was about 8% higher than a simulated "random positioning" method; the resampling process had an additional 6% performance improvement. The performance of a deep learning algorithm for detecting oral cancer can be enhanced by capturing oral photos centered on the lesion. Primary oral cancer diagnosis using smartphone-based images with deep learning offers promising potential.

The aviation industry is constantly changing and to keep up with the trends of air passengers we need predictive models. In this paper, we explore the use of Information Fusion methodologies and classical time series techniques to forecast how many passengers will be traveling by air. Predicting passenger demands is a task, due to various factors that influence travel patterns. The existing models often struggle to capture the dynamics in this field so it's crucial to develop accurate forecasting methods. By leveraging information fusion techniques like smoothing and Autoregressive Integrated Moving Average (ARIMA) our research creates models based on historical data of air passenger volumes. These techniques combine machine learning algorithms and time series analysis to identify dependencies and patterns in the dataset. Through evaluations and comparative analyses, our proposed models demonstrate promising capabilities in forecasting future air passenger volumes. Proof-of-concept experiments based on 5-fold cross-validation demonstrate the efficacy of the proposed approach in capturing underlying trends and seasonality within the dataset.

Diagnosing Parkinson's Disease (PD) can be quite challenging as it presents with symptoms and lacks biomarkers. Nevertheless, the use of data fusion, which combines types of data using machine learning techniques holds promise, for the timely detection of the disease. In this study, we explore the application of data fusion by employing Principal Component Analysis (PCA) as a step to reduce complexity. We then utilize the K Nearest Neighbors (KNN) classification to improve the accuracy of PD diagnosis. By analyzing nonlinear features associated with PD from a dataset PCA helps us extract attributes while maintaining important variations in the data. Subsequently, KNN is employed to identify patterns in this reduced feature space and effectively distinguish between individuals with PD and those who are healthy. Our results show improvements when using the KNN classifier compared to state-of-the-art approaches. This demonstrates its effectiveness in detecting PD leading to promising outcomes and providing a framework for precise PD diagnosis.

Survival analysis remains an important area in predictive modeling, especially in cases where event timing information is critical.  This work presents a research effort to investigate the application of LightGBM, a high-performance high-throughput model, to conduct an improved fusion of decisions from multiple trees to reach survival analysis. Our objective is to address the challenge of developing correct predictive models while advancing computational effectiveness.  Based on a case study of live disaster scenarios, the proposed approach applies and compares LightGBM with traditional prediction methods, which involve careful design engineering, and model training with LightGBM tree structure refinement. The results obtained from fair experimentation and comprehensive predictive performance evaluation demonstrate the robustness of LightGBM in increasing the accuracy of relevant classification tasks toward survival analysis. Furthermore, the findings highlighted that the combination of excellent tree depth for cutting and multi-thread optimization promotes efficient computational complexity and prediction accuracy.

This research focuses on the identification of passengers, in dimensions using information fusion as a tool. We recognize the challenges involved in identifying individuals who have been transferred to alternate dimensions and in this study we make use of CatBoost, an open source machine learning algorithm to address this problem. Our approach includes a preprocessing strategy that involves filling in missing values using techniques like priori distribution terms, which helps ensure the reliability of our dataset. By leveraging CatBoosts ability to handle variables and prevent overfitting we achieve results in accurately predicting passenger movement across dimensions. Our analysis highlights CatBoosts effectiveness in identifying patterns within data leading to more precise predictions for interdimensional passenger transportation. Additionally we incorporate techniques, like Greedy TS augmentation to enhance the adaptability of the algorithm and improve precision while reducing bias in modeling. Proof-of-concept experiments demonstrate that the proposed fusion system not only advances predictive modeling in niche domains but also paves the way for broader applications of machine learning in deciphering complex phenomena beyond traditional realms, marking a significant stride in understanding and addressing unconventional challenges.

The Internet of Medical Things (IoMT) revolutionizes healthcare, enhances patient care, and optimizes workflows. However, the integration of IoMT introduces concerns related to privacy and security. In addressing these issues and aiming to bolster privacy and data security, this study presents a novel cybersecurity framework based on blockchain (BC) technology. The primary goal is to ensure secure communication among IoMT devices, preventing unauthorized access and tampering with sensitive data. The proposed framework is implemented in a model designed for classifying electrocardiogram (ECG) signals, utilizing two datasets: a Medical Technology Database (MTDB) with a limited sample size and the Massachusetts Institute of Technology–Beth Israel Hospital (MITBIH) dataset with a more extensive sample size. The datasets are subsequently partitioned into training and testing data. Feature extraction and selection are performed using the Pan-Tomkins and genetic algorithms. To enhance security, BC technology is employed to encrypt the test data. Finally, signal classification is performed using the support vector machine (SVM) classifier. Thus, the model trained on the MITBIH dataset outperforms its small data counterpart, achieving an impressive accuracy rate of 99.9%. Additionally, the model exhibits a true positive rate (TPR) and true negative rate (TNR) of 100%, an F-score of 100%, and a positive predictive value (PPV) of 100%.

The COVID-19 pandemic necessitated a swift shift to online learning, affecting students differently. We investigated the experiences of 62 students with disabilities in this new educational landscape. Online learning tools raise concerns about privacy and security, making it crucial to explore students' perceptions in these areas. Our findings reveal that while students with learning disabilities appreciate online learning's flexibility, they need more guidance and support. Neurodiverse students with learning disabilities are particularly aware of the need for a secure online learning environment. These insights underscore the unique educational needs of students with disabilities in online education. In Personal Records, authenticating individuals, especially those with visual impairments, is critical. Our research combines education with cutting-edge technologies, like blockchain and machine learning, to enhance biometric authentication for visually impaired individuals. Proposed work focuses the Highly Secure Blockchain-Based Compressive Sensing (HSBCS) system, which uses blockchain for data integrity and machine learning for secure biometric authentication. Our research focuses on education and includes comprehensive testing and performance assessments. Results highlight the educational value of the HSBCS system for Students, as it significantly improves Personal Records data security and accessibility. In conclusion, our research offers an innovative, secure solution for biometric authentication in Personal Records, with a strong emphasis on education. It empowers Students to access their student information securely and independently, while enhancing education on data security and integrity. This study underscores the importance of integrating emerging technologies into Personal Records to provide better experiences for Students and address their unique educational needs.

Constraints perceived in different socioeconomic situations reinforce land use patterns and land cover (LULC) at different levels. However, the statistical information regarding the LULC variations encounters enormous significance for the execution and modelling of appropriate environmental variations and resource management with the available remote sensed data from diverse satellite images and advanced computing technologies; information is generally retrieved from the image classification approaches. However, a broader quantitative analysis of various classification approaches is crucial to choosing an effectual classifier model to acquire appropriate land use regions. We concentrate on the Karavetti region and its related fields in this study. We use a Non-Linear Recurrent Convolutional Neural Network (NLR-CNN) to analyze the data statistically. Well-known techniques such as Support Vector Machine (SVM), Random Forest (RF), and Decision Tree (DT), among others are used to evaluate the model performance. High-resolution images and the data points supplied are also used to assess the accuracy of the categorization and prediction. A confusion matrix is generated where the land cover regions show superior classification accuracy with the fusion model. Also, the NDVI facts and additional metrics like loss, error rate and kappa coefficients are analyzed. Therefore, the outcomes show that the anticipated is considered more robust with better performance to enhance the classification accuracy with the specific land cover regions.

This study addresses the burgeoning challenges in autonomous Maritime navigation by employing information fusion methodologies to assess and manage multifaceted risks. The proliferation of autonomous maritime systems has led to a complex interplay among maritime-related, shore-based remote control, environmental, and emergency management factors, necessitating a comprehensive risk evaluation framework. Leveraging a multi-criteria decision-making approach and employing the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), this research presents a methodical analysis of the coupling coordination degree among these risk variables. Through a meticulous examination of historical accident data and information fusion techniques, our study reveals dynamic trends in the comprehensive risk evaluation index, showcasing the evolving nature of risks inherent in autonomous Maritime navigation. The predictive insights gleaned from these analyses forecast an initial increase followed by a peak in accidents, underscoring the urgency for proactive risk mitigation strategies. This study's conclusions emphasize the pivotal role of information fusion methodologies in comprehensively assessing, understanding, and managing risks within autonomous Maritime navigation.

Effective procurement of clinical devices in healthcare demands a sophisticated decision-making approach integrating diverse data sources from multiple devices, brands, and suppliers, particularly within the context of information fusion. This study addresses this challenge by proposing an improved best-worst method harmonized with information fusion techniques and multi-criteria decision-making methodologies. The background emphasizes the dynamic nature of healthcare procurement, necessitating systematic strategies for navigating the complexities of device selection and integration. Recognizing the intricacies inherent in this challenge, the problem statement revolves around enhancing the best-worst method to amalgamate data from clinical devices while concurrently evaluating brands and suppliers. This aims to optimize performance and minimize costs within the information fusion paradigm. Our proposed methodology introduces an augmented best-worst approach, encompassing weighted criteria assessment for clinical devices, brands, and suppliers, providing a more adaptable and nuanced decision-making framework tailored to the information fusion landscape. The results showcase a structured evaluation matrix derived from refined weighted criteria, elucidating the relative performance and strengths across various entities within the healthcare procurement ecosystem. Emphasizing reliability, compatibility, innovation, and quality assurance, this process highlights pivotal factors influencing procurement decisions within the realm of information fusion.

Energy policy implementation relies heavily on assessing savings from retrofitting for energy efficiency. Because of their unique purpose, hospitals need energy-efficient renovations to improve indoor air quality and create a pleasant space for staff and visitors. Because of this crucial distinction, investors' preferences must be considered when deciding on refurbishment plans. Considering elements including energy savings, financial viability, and thermal comfort, this research provides a multi-criteria decision-making (MCDM) approach to guide investors in choosing the most effective remodeling plan for hospital wards. We used the MABAC method as an MCDM fusion method to combine the various criteria and alternatives to select the best one. We used ten criteria and ten alternatives in this study. We compute the weights of criteria to rank the criteria. Then, we used the MABAC fusion to rank the alternatives. The results show the financial viability has the least weight and the building envelope has the highest. We conducted a sensitivity analysis to show the stability of the results in this study.

Groundwater recharge is essential in establishing reliable groundwater supplies in a region. Groundwater is a vital natural water resource, but its quantity and quality may vary significantly from one area to another. Growing urbanization and population increase have put a significant demand on groundwater supplies. Using Multi-Criteria Decision-Making (MCDM), several studies have identified good areas for recharging groundwater supplies. To help choose between several types of artificial recharge (AR) structures, we have developed an MCDM approach for this research. We used an MCDM fusion methodology to combine various AR criteria with the alternatives. This study collected eight criteria and eight alternatives. We used the average method to compute the weights of the criteria. Then, we used the COCOSO method as an MCDM fusion method to rank the alternatives. The results show that hydrological conditions are the best criteria, and stakeholder engagement is the lowest weight. The sensitivity analysis is performed to show the stability of the results in this study. 

Even though the transmission and processing speeds of electronic documents have been vastly enhanced, electronic document information may be revealed, counterfeited, tampered with, or otherwise compromised. To maintain corporate success in the marketplace, network security should be essential to the protection of electronic documents. As a result, there is a rising demand for authentication and verification procedures for a variety of important documents, including those used in banking, government, and other transactions as well as certificates and other academic credentials. In recent years, there has been a fast growth of digital watermarking technology, which involves embedding invisible or hidden digital signatures into data without compromising the data's authenticity. Hence, in this paper, we utilize the watermarking technology in the encrypted data using dynamic wavelet transform algorithm to make a document more protected. Now the protected data is sent to cloud database for storage. Integrated digital signature algorithm (SHA-256 + DSA) is proposed in this research to generate digital signature for each document. When recipients download the data, the data is verified for its integrity after extracting the digital signature and encrypted data. This strategy improves record security. We also compare the suggested technique to standard practices and assess its performance based on a variety of indicators to demonstrate its effectiveness.  

Suicide is a significant issue for public health worldwide since suicide is not something that happens randomly but is influenced by social and environmental variables as well. At the same time, effective early diagnosis and treatment may lead to several positive health and behavioural results. Suicide persists undiagnosed and untreated for many reasons, including denial of sickness and cultural and social disgrace. Through the ubiquity of social media, by expressing opinions, thoughts and everyday struggles with mental health on social media, millions of people are sharing their online identity. As opposed to typical retrospective research that uses self-reported surveys and questionnaires, this study assesses the validity of identifying suicidal symptoms using Twitter tweets that were gathered over more than a year, using a variety of online web-blogging sites as points of reference. For recognizing tweets expressing suicidal thoughts, three sets of characteristics are employed for training the dataset employing base and ensemble classifiers. The Rotation Forest (RF) approach is the preferred baseline, and the Maximum Probability Voting Decision approach is used in seven different labelled classes relating to suicide communication and class demonstrating suicidal thoughts. With the suicidal ideation class scoring 0.76 and the suicidal contents for all seven classes scoring 0.82, this revised model was able to attain an F-measure. To increase awareness of the vocabulary made use of on Twitter to express suicidal thoughts, the findings are summarized by highlighting the predictive principal component of suicide communication in classrooms.