Fusion: Practice and Applications

Leukemia is a form of blood cancer that targets white blood cells (WBC) and stands as a major cause of mortality worldwide. During the center of human bones, leukaemia is presented and provides blood cell generation with leukocytes and WBC, and if some cell comes to be blasted, then it grows a fatal illness. For that reason, the analysis of leukaemia in its initial stages aids significantly in the treatment accompanied by saving the life. At present, leukemia analysis is done by visual assessment of biomedical images of blood cells, which is time-consuming, tedious, and wants to train specialists. Consequently, the lack of an early, automatic, and effectual leukemia recognition model is a major problem in hospitals. A few automated techniques like deep learning (DL) and Machine learning (ML) methodologies for leukemia cancer identification are presented that offer remarkable and effectual results. This study develops a Robust Multimodal Fusion of Transfer Learning Framework for Leukemia Cancer Detection and Classification (RMFTLF-LCDC) algorithm. The RMFTLF-LCDC system mostly suggests identifying and classifying the existence of leukemia cancer on biomedical imaging. At first, the RMFTLF-LCDC model applies image preprocessing using a kernel correlation filter (KCF) to eliminate the noise. For the feature extraction process, the multimodal fusion of CapsNet models, including RES-CapsNet, VGG-CapsNet, and GN-CapsNet are implemented to improve the representation of features by providing more accurate initial information to subsequent capsule layers. In addition, the recurrent spiking neural network with the spiking convolutional block attention module (RSNN-CBAM) technique is performed for the leukemia cancer detection process. At last, the improved Harris hawk optimization (IHHO) approach-based hyperparameter choice can be executed to improve the classification outcomes of the RSNN-CBAM system. The efficiency of the RMFTLF-LCDC method has been validated by comprehensive studies using the benchmark image dataset. The numerical result shows that the RMFTLF-LCDC method has better performance and scalability across other recent techniques.

Spam e-mail has become a pervasive nuisance in today's digital world, posing significant challenges to efficient communication and information dissemination. Dealing with huge amounts of data with irrelevant and redundant features, which leads to high dimension. Nowadays, with the growth of using the internet, finding the secure E-mail classification system for cloud computing is a very important topic. Additionally, determining the best algorithm for choosing a subset of features has a big impact on how well automatic email classification works, making it one of the major issues. Among these is the Differential Evolution (DE) algorithm, which is computationally costly because of the slow convergence rate and evolutionary process. To address these issues, this study offers an intelligent scheme called Opposition Differential Evolution (ODE), which combines the Opposition Based Learning (OBL) and DE algorithms for effective automated feature subset selection. Its effectiveness is assessed using the support vector machine (SVM) to present a strong performance when evaluating the e-mail spam classification rate. Moreover, the OBL is used to accelerate and increase the convergence rate of traditional DE. To determine which features, contribute most to the reliability of the email spam classification, the subset features based on ODE that was used as feature subset selection are used.To assess the effectiveness of the proposed scheme, extensive experiments are conducted on spambase” and “spamassassin” benchmark email datasets, comprising a diverse collection of spam and non-spam emails. The results demonstrate that the Opposition Differential Evolution (ODE) algorithm yields superior performance compared to traditional machine learning and evolutionary techniques, displaying its robustness and efficiency in identifying spam emails accurately. The ODE algorithm effectively handles high-dimensional feature spaces, enhancing the model's discriminatory power while maintaining computational efficiency. Compared to the suggested ODE-SVM technique, which yields a result of 96.79 percent, the full-feature accuracy result was 93.55 percent. Additionally, empirical results demonstrate that our scheme may efficiently increase the number of features needed to improve the accuracy of the email spam classification.

Steganalysis can be defined as the science that addresses the process of identifying and detecting hidden information or data within various types of digital media. Recently, Deep Learning (DL) approaches have been employed to build steganalysis systems. However, the problem with steganalysis systems adopting a DL approach is their low accuracy and their need for effective datasets to be used for the training. In this paper, we introduce a DL-based Steganalysis system for the detection and classification of hidden content in images. Our system, called Steg-Analysis Convolutional Neural Network (SA-CNN), relies on a Convolutional Neural Network (CNN) and uses High Pass Filter (HPF) and extra-embedded data. We also propose a preprocessing-based data hiding method to increase the accuracy of SA-CNN in detecting hidden content. Therefore, this ensures the imperceptibility of images used for training SA-CNN. In addition, we use another CNN, called Malicious-Benign Classification CNN (MBC-CNN), that we have developed to classify the extracted hidden content into Malicious or Benign classes. Compared with existing systems, SA-CNN shows a better performance in terms of accuracy, under increased hiding rates ranging from 0.1 to 1.0 bpp, reaching 90%.

An inventive deep learning-based method for identifying financial fraud, revolutionizing e-commerce security in the process. The research offers a state-of-the-art setup that makes use of deep learning computations in the dynamic world of online exchanges, where the possibility of fraudulent activity is a danger. Since frauds are known to be erratic and lack consistency, it might be challenging to spot them. Con artists exploit the latest developments in technology. They manage to evade security measures, which results in millions of dollars being lost. One method of tracking fraudulent exchanges is to use information-mining techniques to investigate and detect unusual behaviours. Interactions. In contrast to deep learning techniques as auto encoders, convolutional neural networks (CNN), restricted Boltzmann machines (RBM), and deep belief networks (DBN), this paper aims to benchmark several machine-learning techniques, such as k-nearest neighbour (KNN), irregular forest, and support vector machines (SVM). The three-evaluation metrics that are really employed are the Area Under the ROC Curve (AUC), the Matthews Correlation Coefficient (MCC), and the Cost of Failure.

Artificial intelligence techniques including deep learning play a major role in all fields and in line with the advancement in technology. Handwritten digit recognition is an important issue in the field of computer vision, which is used in wide applications such as optical character recognition and handwritten digits. In the current research, we describe a unique deep learning technique that uses a Convolutional Neural Network (CNN) framework with better normalization algorithms and adjusted hyperparameters for improved efficiency as well as generalize. Contrasting conventional techniques, our methodology concentrates on minimizing overfitting through the use of adjustable rate of abandonment and innovative pooling procedures, resulting in greater accuracy in handwriting number classification. Following considerable research, the recommended approach obtains an outstanding classification accuracy of 99.03%, proving its ability to recognize intricate structures in handwritten numbers. The approach's usefulness is reinforced by a complete review of measures including recall, accuracy, F1 score, as well as confuse matrix assessment, which show improvements throughout all digit categories. . The results of the investigation highlight the innovative conceptual layout and optimization methodologies used, representing a substantial leap in the realm of number identification.

Evaporation plays a significant role in managing water resources and is an important indicator in risk and crisis management, particularly in operating reservoirs and dams. Precise predictions of evaporation rates are crucial to effective water resource management, and various modelling methods, including AI and autoregression, have been employed to create accurate models. This makes it more important to use innovative technology to continuously monitor this phenomenon with accurate scientific results, allowing decision-makers to be aware of and prepare for potential drought risks and crises. In this study, therefore, we propose the establishment of a mechanism that would include analyzing and exploring the data used in this study (Evaporation) and cleaning up the impurities of actual and lost values to obtain accurate data that would serve as actual inputs to ARIMA model that will adopt in this study, This mechanism would contribute to the performance and efficiency of this model using time series data to accurately predict future trends of evaporation plants in the water of the Mosul dam. Our objective is to explain the diversity of climate policies and actions using a data-based approach to analyzing integrated parameters over the years, etc. This is complemented in depth by how different methods of extracting data behaviour are used to study model forecasts. This collaborative study aims to enhance future studies by using more comprehensive datasets with more learning models. The researchers believe in the power of sharing knowledge and are thus committed to sharing the results of other causes outside of global warming that contribute to climate change.

This study aimed to measure the impact of using "Box-to-Box" technology in improving physical and technical abilities for football players under 19 years old at Najma Sinai Sports Club, North Sinia the research highlights the global appeal of football but also offers insight into how advancements in training can help to improve player performance, some teams tend to cling old-school tactics which undermine progress. The study evaluated a 12-week "Box-to-Box" training program using an experimental design with pre and post intervention measurements for 23 players. The results showed that while agility, endurance, speed, and muscle strength test scores significantly improved; passing accuracy and dribbling efficiency were also enhanced during composite skill performance. These findings reaffirm that "Box-to-Box" Training is the way to go for developing key competencies and improving performance, in general. The study suggests including this new technology in traditional training routines, asserting that it has now become essential for player assessment and improvement. It also proposes a wider perspective on the long-term use of "Box-to-Box" technology in different populations and sports, as well as new functional training for specific football positions.

Recently, it has been observed that the tourism industry is undergoing a fundamental change due to the rapid development of virtual tour technologies, especially artificial intelligence. This paper therefore aims to provide an overview of this new development from the early 2000s to the current environment in global tourism. We present, in a historical context, the main developments and applications of virtual tours and AI through a systematic review of literature, industry reports and empirical data from different sectors of the tourism industry. Our findings suggest that the adoption of the technologies under review, enhanced by data fusion, has significantly reshaped the way tourism experiences are conceptualized, delivered, and consumed. Data fusion combines information from multiple sources, enabling richer insights and a more comprehensive understanding of traveller behaviours and preferences. While virtual tours have emerged as a powerful tool for destination marketing, cultural preservation, and accessibility, AI, combined with data fusion, has also transformed the landscape by enabling more personalized travel planning, responsive customer service, and data-driven decision-making. This integration allows tourism providers to create seamless and engaging experiences tailored to individual needs, making tourism more accessible and efficient. In each case, these innovations have raised important questions about authenticity, sustainability, and the future of traditional tourism business models. We will present a critical comparison of virtual and physical tourism experiences in different regions and market segments, providing insights into the interplay of technological innovation, economic imperatives, and socio-cultural dynamics in the digital age. We conclude by reflecting on the implications for post-pandemic recovery, responsible tourism and global cultural exchange through virtual tours and AI. The findings of the study add to the growing body of knowledge on the digitalization of tourism and provide useful insights for practitioners, policy makers and researchers interested in the rapidly changing landscape of this industry.

Electric vehicles (EVs) have gained significant traction due to their environmental benefits and potential to revolutionize the transportation sector. Integrating EVs into the Vehicle-to-Grid (V2G) network presents an innovative solution for optimizing energy transactions and grid stability. However, managing energy transactions during peak hours poses a challenge. This research proposes a novel approach that combines the Deep Q-Network (DQN) algorithm with block chain technology to enhance energy transactions in the V2G network. In this study, a V2G network model is introduced consisting of EVs, charging stations, a grid control center, and a block chain infrastructure. The block chain ensures transparency, security, and decentralized energy transactions. The DQN algorithm learns optimal action policies based on current states and rewards, contributing to grid stability. To incentivize EV owners for peak-hour energy contributions, a block chain-enabled rewarding mechanism is implemented. The proposed methodology is rigorously evaluated through simulations conducted in a custom environment that emulates V2G network dynamics. Performance metrics such as load shifting efficiency, peak demand reduction, and energy efficiency are employed for comprehensive assessment. The proposed method showcases superior performance compared to traditional load shifting and demand response strategies. Furthermore, comparative analyses are conducted against different state-of-the-art methods, demonstrating the effectiveness of our approach. The results underscore the potential of integrating DQN-based energy management with block chain technology to achieve grid stability and incentivize sustainable energy behaviors. This research contributes to the advancement of smart grid technologies, paving the way for a more sustainable and efficient energy ecosystem.

Retinopathy is a progressive and common retinal disease that most progressive diabetics suffer from and causes blood vessels in the retina to swell and leak blood and fluid. This condition requires timely diagnosis via medical experts to prevent causing visual loss among patients. To enhance the feasibility of checking many persons, diverse deep-learning schemes have recently been developed for diabetic retinopathy detection. In this paper, retinopathy image detection system based on diverse deep learning schemes (VGG-19, DenseNet-121, and EfficientNet-B6) has been presented. The implemented deep learning schemes with multi-label classification are trained and tested using the Asia Pacific Tele Ophthalmology Society (APTOS-2019) dataset, and the two combined datasets Indian Diabetic Retinopathy Image Dataset (IDRiD) and Messidor-2. The system outcomes of classification are exhibited as sensitivity, precision, F1Score, and accuracy measurements, and the system performance is compared with recently existing related systems. The attained outcomes indicate that the implemented EfficientNetB6 network outperforms peers’ schemes and related systems via realizing supreme accuracy using balanced multi-class retinopathy datasets.

This review provides an in-depth exploration of machine learning (ML) applications in healthcare, focusing specifically on predictive models for COVID-19 transmission among vaccinated individuals. It underscores the pivotal role of ML in disease forecasting and prognosis, showcasing its potential to enhance healthcare outcomes in pandemic contexts. Key challenges of COVID-19, such as the high transmission rate of asymptomatic carriers and the effectiveness of containment strategies, are analyzed to highlight areas where ML can offer significant advantages. The study aims to develop an advanced forecasting model for COVID-19 transmission using diverse supervised ML regression techniques, including linear regression, LASSO, support vector machine, and exponential smoothing, applied to an extensive COVID-19 patient dataset. The insights generated from this review support efforts to combat COVID-19 and improve public health strategies, demonstrating ML's vital contribution to pandemic management and healthcare resilience.

Solving the video compression problem requires a multi-faceted approach, balancing quality, efficiency, and computational demands. By leveraging advancements in technology and adapting to the evolving needs of video applications, it is possible to develop compression methods that meet the challenges of the present and future digital landscape. To address these objectives, machine learning and AI approaches can be utilized to predict and remove redundancies more effectively, optimizing compression algorithms dynamically based on content. Still, state-of-the art neural network-based video compression models need large and diverse datasets to generalize well across different types of video content. Wavelets can provide both time (spatial) and frequency localization, making them highly effective for video compression. This dual localization allows wavelet transforms to handle both rapid changes in video content and slow-moving scenes efficiently, leading to better compression ratios. Yet, some wavelet coefficients may be more critical for maintaining visual quality than others. Inaccurate quantization can lead to noticeable degradation. For the first time, the suggested model combine Quantum Wavelet Transform (QWT) and Neural Networks (NN) for video compression. This fusion model aims to achieve higher compression ratios, maintain video quality, and reduce computational complexity by utilizing QWT’s efficient data representation and NN’s powerful pattern recognition and predictive capabilities. Quantum bits (qubits) can encode large amounts of information in their quantum states, enabling more efficient data representation. This is especially useful for encoding large video files. Furthermore, quantum entanglement allows for correlated data representation across qubits, which can be exploited to capture intricate details and redundancies in video data more effectively than classical methods. The experimental results reveal that QWT achieves a compression ratio of almost twice that of traditional WT for the same video, maintaining superior visual quality due to more efficient redundancy elimination.

In the realm of education, understanding the impact of different teaching styles on student engagement and satisfaction is essential. Recent advancements in sentiment analysis provide new avenues for evaluating student feedback, particularly through informal channels such as social media. While formal student evaluations offer structured feedback on teaching styles, they may not fully capture the nuanced opinions and sentiments expressed by students in informal settings, such as social media. This research aims to address the gap by integrating sentiment analysis of social media data to evaluate teaching effectiveness across various styles and comparing it with formal evaluation results. This study employs sentiment analysis using the VADER (Valence Aware Dictionary and sEntiment Reasoner) tool to analyze student posts on social media platforms. The analysis includes the extraction of sentiment distributions, identification of common keywords, and tracking of sentiment trends over time. Additionally, formal student evaluations (Likert scale) are collected to offer a direct comparison. The teaching styles analyzed include lecture-based teaching, project-based learning, flipped classrooms, online learning, hybrid learning, and traditional exam-based learning. The findings demonstrate that student sentiment varies significantly across teaching styles. Flipped classrooms and project-based learning received the highest positive sentiment scores, while traditional exam-based teaching showed the most negative sentiment. Social media feedback tended to align with formal evaluations for certain teaching styles, such as the flipped classroom and hybrid learning but showed divergence in others, like online learning, which received higher sentiment in social media feedback. Trends over time reveal evolving sentiments, with fluctuating satisfaction as the academic semester progressed. The integration of social media sentiment analysis provides a more dynamic and real-time understanding of student experiences, offering deeper insights into teaching style effectiveness.

Epilepsy is a neural condition that is rather prevalent and affects a sizeable portion of the average population all over the world. Throughout its history, the illness has constantly be located of significant status in the pitch of biomedicine due to the dangers it poses to people's health. Electroencephalogram (EEG) recordings are a method that may be utilized to evaluate epilepsy, which is defined by the occurrence of seizures that occur repeatedly and without any apparent cause. Electroencephalography, often known as EEG, is a method that is utilized to assess the electric movement located within the brain. The examination of electroencephalogram data is an essential component in the field of epilepsy research, since it allows for the early detection of epileptic episodes. On the other hand, the generation of models that are independent of individual characteristics is a significant challenge. Extensive efforts have been directed to the creation of classifiers that are tailored to specific patients. In this thesis, the cross-patient viewpoint is the primary focus of investigation; nevertheless, the heterogeneity of EEG patterns among people presents a challenge to this investigation. An examination of the similarities and differences of the pattern recognition algorithms that are applied for the diagnosis of epileptic episodes based on EEG data was taken. SVM (Support Vector Machine) and KNN (K-Nearest Neighbor) were the approaches that were under consideration for evaluation. According to the findings of our analysis, the two approaches exhibit comparable levels of performance; however, KNN attained a slightly greater level of accuracy in some situations on occasion.

This paper proposes a better solution for EEG-based brain language signals classification, it is using machine learning and optimization algorithms. This project aims to replace the brain signal classification for language processing tasks by achieving the higher accuracy and speed process. Features extraction is performed using a modified Discrete Wavelet Transform (DWT) in this study which increases the capability of capturing signal characteristics appropriately by decomposing EEG signals into significant frequency components. A Gray Wolf Optimization (GWO) algorithm method is applied to improve the results and select the optimal features which achieves more accurate results by selecting impactful features with maximum relevance while minimizing redundancy. This optimization process improves the performance of the classification model in general. In case of classification, the Support Vector Machine (SVM) and Neural Network (NN) hybrid model is presented. This combines an SVM classifier's capacity to manage functions in high dimensional space, as well as a neural network capacity to learn non-linearly with its feature (pattern learning). The model was trained and tested on an EEG dataset and performed a classification accuracy of 97%, indicating the robustness and efficacy of our method. The results indicate that this improved classifier is able to be used in brain–computer interface systems and neurologic evaluations. The combination of machine learning and optimization techniques has established this paradigm as a highly effective way to pursue further research in EEG signal processing for brain language recognition.