Fusion: Practice and Applications

In e-commerce, web mining for page recommendations is widely used but often fails to meet user needs. To address this, we propose a novel solution combining semantic web mining with BP neural networks. We process user search logs to extract five key features: content priority, time spent, user feedback (both explicit and implicit), recommendation semantics, and input deviation. These features are then fed into a BP neural network to classify and prioritize web pages. The prioritized pages are recommended to users. Using book sales pages for testing, our results demonstrate that this solution can quickly and accurately identify the pages users need. Our approach ensures that recommendations are more relevant and tailored to individual preferences, enhancing the online shopping experience. By leveraging advanced semantic analysis and neural network techniques, we bridge the gap between user expectations and actual recommendations. This innovative method not only improves accuracy but also speeds up the recommendation process, making it a valuable tool for e-commerce platforms aiming to boost user satisfaction and engagement. Additionally, our system’s ability to handle large datasets and provide real-time recommendations makes it a scalable and efficient solution for modern e-commerce challenges.

Machine learning (ML) is the most up-to-date approach for classifying music genres. Due to technological ML advancements, its technologies can help in music genre recognition best. In machine learning, effective fusion of different features could improve recognition performance. Hence, this paper presents a new robust method for Arabic music classification based on the fusion of different sets of features. Frequency-domain, time-domain, and cepstral domain features have been combined and compared with other state-of-the-art approaches. Four machine-learning models that categorize music into its appropriate genre have been created: support vector machines (SVM), K-nearest neighbors (KNN), naïve Bayes (NB), and random forest (RF) classifiers were utilized in a comparative analysis of other ML algorithms, and the accuracy of these models has been assessed and derives the appropriate conclusions. To assess the performance of our method, two various datasets are used: the collected dataset, namely Zekrayati, which was collected by authors in favor of this paper, and the global GTZAN dataset, which was used to compare with previous studies. The experimental findings indicated that the SVM exhibited a higher optimal accuracy of 99.2% and has proven that the fusion proposed features will help to classify music in different fields.

This study proposes an intelligent system designed to detect and manage epidemic outbreaks within institutional settings by leveraging a fusion of advanced AI technologies. The system operates through five key stages: symptom-based diagnostic testing, AI-powered cough detection, analysis of X-ray and CT scan images using Convolutional Neural Networks (CNN), evaluation of vital signs, and the geolocation of COVID-19 patients using GPS. Cough detection is enhanced by integrating Short-Time Fourier Transform (STFT) and Mel-Frequency Cepstral Coefficients (MFCC). Trained on an extensive dataset comprising over 5,856 CT scans, 7135 X-ray images, and over 30,000 crowdsourced cough recordings, the system demonstrates a high accuracy rate of 95% in identifying potential epidemic cases. This fusion of techniques offers a robust solution for early detection and rapid intervention, significantly mitigating the risk of widespread transmission within high-density environments.

Face detection is important in computer vision and image processing, particularly in surveillance, security systems, video analytics, and facial recognition applications. However, face detection algorithms face challenges like position variations, lighting fluctuations, size and resolution differences, facial expressions, and background clutter. This research aims to develop a system that achieves high accuracy in detecting and localizing faces using local descriptors and spatial feature extraction techniques, specifically the Histogram of Oriented Gradients method (HOG). Using videos from the YouTube Face database, features were extracted from frames and trained using a convolutional neural network (CNN). The HOG technique achieved a 94% accuracy rate and good localization compared to CNN without feature extraction.

In Wireless Sensor Networks (WSN), congestion control plays a crucial role as the traffic load surpasses the capacity of each major channel. The WSN constrained resources must be taken in consideration while devising such strategies to get the best throughput. Various factors are contributed in the congestion; the primary factor is the over flowing buffer, packet loss, reduce network throughput and loss of energy. This research, studies path load distribution in novel networks, including anonymous communication. Initially there is a chance that the public Wi-current Fi approach will result in notable imbalances. We next modify an optimal path-selection algorithm and use flow level visualization to show that this results in a substantially improved network load balance. Web-based Congestion Control (WCC) needs to make it possible to give WCC channel flows a distinct quality of service (QoS) in order to overcome this difficulty.

The increasing use of credit cards, especially for online payments, has led to a significant increase in fraud involving credit card payment technologies. Financial companies must enhance fraud detection systems to mitigate significant losses. This study introduces a methodology for developing a credit card fraud detection system that uses the Synthetic Minority Oversampling Technique (SMOTE) to address an imbalanced dataset problem and an attention layer to identify important features in the input sequence, two long short-term memory (LSTM) layers modeling long-run dependencies within a sequence of transactions, a dropout layer that neglects values lower than 0.3, and two dense layers, which allows enhancing the accuracy of prediction of fraudulent transactions. When implemented, the proposed system achieves an accuracy of 0.9434% on the IEEE dataset, 0.9850% on the Banksim dataset, and 0.9757% on the European dataset. This methodology shows improvements in fraud detection, emphasizing its ability to enhance financial security systems and reduce misclassification in credit card transactions.

The stock price exhibits quick and extremely nonlinear fluctuations in the financial market. A prominent worry among scholars and investors is the correct prediction of short-term stock prices and the corresponding upward and downward trends. Financial organizations have successfully incorporated machine learning and deep learning techniques to anticipate time series data accurately. Nevertheless, the precision of these models' predictions still needs improvement. Most current studies employ single prediction algorithms that cannot overcome intrinsic limitations. This paper proposes a methodology that utilizes the MUTUAL, principal component analysis (PCA), and Long Short-Term Memory (LSTM) model to accurately simulate and predict the variations in stock prices. The technology is utilized for the three global stock market datasets: TSLA, S&P500, and NASDAQ. The highest level of improvement achieved is a correlation of 99%. Furthermore, there is a reduction in error for the metrics MSE, MAPE, and RMSE, with improvements of 0.0001, 0.009, and 0.01 correspondingly.

This study presents a novel approach to predictive modeling of muscular performance and fitness progression using artificial intelligence techniques. Leveraging advanced machine learning algorithms, including artificial neural networks (ANN), support vector machines (SVM), and gradient boosting machines (GBM), we develop a comprehensive model capable of accurately forecasting key metrics related to muscular strength, endurance, and overall fitness. Extensive experimentation and evaluation demonstrate the superiority of the proposed method over existing algorithms across a range of performance metrics, including accuracy, precision, recall, F1-score, and error metrics such as mean squared error (MSE), root mean squared error (RMSE), and mean absolute error (MAE). Our findings highlight the importance of feature selection techniques and model hyperparameter optimization in driving predictive performance, underscoring the need for careful model development and tuning. The practical implications of our research extend to sports science and athletic training, where the proposed method can inform personalized training strategies tailored to individual athletes' needs and goals. Moving forward, further research is needed to validate the robustness and generalizability of the proposed method across different populations and athletic disciplines, as well as to explore its integration with real-time data sources for more dynamic and responsive training programs.

This article focuses on improving the accuracy and efficiency of multimodal human motion analysis using advanced techniques. Initially, Generative Adversarial Networks (GANs) were used for skeletal enhancement, and then Contrast-Limited Adaptive Histogram Equalization (CLAHE) was applied on the enhanced images to check the quality Joint-level. Limb-level, Temporal, Statistical Features are effectively recovered from contrast enhancing images. Furthermore, with the selected optimal features acquired from PutterFish Customized Serval Optimizer (PFCSO), the RehabNet++ architecture that makes the human movement assessment has been trained. This PFCSO model has been developed based on the inspiration acquired from the Pufferfish Optimization Algorithm (POA) and the Serval Optimization algorithm (SOA), respectively. The RehabNet++ architecture includes an optimized Multilayer Perceptron (O-MLP), STR-ResNet architecture, Attention-based Convolutional Neural Networks and Transfer Learning. The O-MLP model has been formulated by optimizing the hidden layers of MLP using the PFCSO model. In addition, Grad-CAM visualization is included to provide a graphical description for model selection. A comparative study has been conducted to test the proposed deep learning algorithm against the original methods using the Kimore dataset. This analysis is implemented in PYTHON and is dedicated to multimodal human motion analysis.

For many years, scientists have studied the way people express their emotions through body language and facial expressions. However, it is extremely difficult to accurately interpret the emotions of a person from just a single image. Interpreting facial emotions in photographs is a complex task. It is challenging to accurately detect facial emotions with the help of neural networks when the face is occluded with fragmentary blocks. With the advent of technology, emotion detection has become more accurate and reliable. It is now possible to use facial expression recognition in images to detect emotions such as happiness, sadness, anger, fear, surprise, and more. This research discusses the effectiveness of using neural networks to identify facial emotions in photographs with occlusions present. The datasets like Fer2013 dataset, CREMA-D and RAVDESS were used to train the model and the datasets were altered by implanting occlusions randomly in the images. The altered datasets were also used to evaluate the model. The challenges and opportunities that arise when neural networks are used in this context are explored. Additionally, insight is also provided into the best approach to accomplish the task.

Sign language recognition is important for enhancing message and user-friendliness for the community of deaf and hearing-impaired people. This paper proposes a Malayalam Sign Language (MSL) method using sign language that emerged from the state of Kerala. The main factor contributing to this emergence of such regional sign language is the absence of a standardized and consistent approach to the use of Indian Sign Language (ISL) in various states. This is due to the variations in signs, grammar, and syntax used in different regions. The system uses the You Only Look Once v8 (YOLOv8) algorithm-based object detection method which is based on Convolution Neural Network (CNN), a widely accepted deep learning neural network design employed mainly in computer vision. As the dataset for MSL is not publicly available, we used an MSL video from YouTube provided by the National Institute of Speech and Hearing for training a custom model. We pre-processed the video to extract the frames and annotate them with sign labels. Then, we trained the YOLOv8 algorithm on the annotated frames to detect the hand region and recognize signs in real time. The proposed approach achieved an accuracy of 97.21% calculated from the mean Average Precision value on the MSL dataset. The result achieved outperformed other existing approaches even while using less dataset count compared to others.

The incidence of cancer cases has been rising rapidly over the last few decades. Skin cancer is one of the widely found types of cancer, is further classified into two main types, Melanoma and Non-Melanoma. Though Melanoma is less common than other types of skin cancer, it can be lethal if not treated promptly. But it is not the only type of skin lesion that needs attention. It becomes necessary to promptly identify and classify the skin lesions for the recovery of the patient. The machine learning models of Deep Learning prove to be very efficient in this regard. Hence, we developed a deep learning model which is an ensemble of InceptionV3, Xception and ResNet152 models. It can classify the skin lesions into seven main types -Melanoma, Melanocytic Nevi, Benign Keratosis-like lesions, Basal cell carcinoma, actinic keratosis, vascular lesions, Dermatofibroma. The method was applied to dermoscopic images from the HAM10000 dataset. The presence of noise and artifacts in the images makes it difficult to classify. So, as a preprocessing step, we performed hair removal on the dermoscopic images which is a series of methods that starts with blackhat filtering, subsequently creating a mask for inpainting and then applying the inpainting algorithm. Further Contrast enhancement was performed by applying the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm on the luminance channel of HSV image to improve the contrast of the image and also makes sure that it is not over-amplified. It is then followed by Skin Lesion Segmentation where a grabcut algorithm is applied on the enhanced image which segments the image. Thus, the segmented images are produced which are fed to the Model for training and testing. To cope up with the unbalanced dermoscopy image dataset available, we performed Image augmentation on the images generated in the previous step which alters the existing images to create some more images for the model training process, thus solving the problem of paucity of dataset and substantially increases the performance of the model. The final dataset generated is fed to the three deep learning models InceptionV3, Xception and Resnet152 which achieved an accuracy of 84.6%, 86.5% and 86.7% respectively. These were later given to two different ensemble models - Stacking and Random Forest. The Stacking model achieved an accuracy of 88.6% and Random Forest achieved an accuracy of 92.59%. The proposed system includes a GUI for a good user experience.

A wireless body area network (WBAN) is a wireless sensor network (WSN) that is essential to monitor patient health. Sensor nodes (SNs) are commonly positioned either inside or outside the patient's body within this network. These nodes have the ability to send data to the sink node if any functional modifications in the patient are observed. Delivering efficient routing and energy management of network nodes is a complex effort in WBAN. The energy efficiency of SNs is a primary challenge to the effective deployment of WBAN. To handle this problem, a new metaheuristic optimization algorithm called Elk Herd Optimizer (EHO) is proposed in this research. This research aims to focus on energy-efficient routing methods in WBAN sensors that are connected to the human body to enhance health monitoring efficiency. The proposed WBAN model includes the deployment of eight biosensor nodes on the human body. The primary objective is to minimize the energy utilization of WBANs by selecting the most appropriate cluster heads (CHs) based on the EHO. The EHO-based routing protocol showed higher performance in WBANs in terms of energy consumption, End-to-End (E2E) delay, packet delivery rate (PDR), network lifetime (NLT), packet loss rate (PLR), and throughput. The research model was validated by comparing its findings with the existing routing protocols. The research model surpassed all the comparable models in terms of energy consumption, latency, NLT, PDR, PLR, and throughput. The routing protocol based on the EHO algorithm improves energy efficiency by effectively selecting CHs and routing paths. The EHO model efficiently reduces the total time delay, which is essential for monitoring health in real time. It achieves a high PDR while maintaining a low packet loss rate. Furthermore, the EHO-based routing extends the longevity of the network. Additionally, it enhances network performance, hence facilitating uninterrupted and dependable monitoring of health data.

With the prevalence of stress-related disorders on the rise, there is an increasing demand for advanced methodologies that can effectively detect and analyze stress levels. In response to this need, this research explores the integration of Fast Fourier Transform (FFT), Support Vector Machine (SVM), and Convolutional Neural Network (CNN) techniques for unlocking insights into stress dynamics from Electroencephalogram (EEG) signals. Stress, a multifaceted phenomenon with far-reaching implications for mental health, necessitates innovative approaches for its identification and management. The study begins by elucidating the complexity of stress and its impact on individuals' well-being, highlighting the urgency for accurate and efficient stress detection methodologies. Building upon this foundation, the technical intricacies of FFT, SVM, and CNN integration are explored, elucidating their respective roles in the stress detection framework. The FFT method is employed for spectral analysis of EEG signals, providing a foundation for identifying stress-related patterns in the frequency domain. The application of Artificial Neural Networks (ANNs) for feature extraction and classification is explored, leveraging their capacity to discern intricate relationships within EEG data structures. Complementing ANNs, Support Vector Machines (SVMs) are harnessed for stress level classification, capitalizing on their robustness and efficiency in handling high-dimensional data spaces. Furthermore, Convolutional Neural Networks (CNNs) are integrated into the framework to automatically learn hierarchical features from raw EEG signals, enhancing the accuracy and efficacy of stress detection methodologies. Through comprehensive evaluation and comparison with existing algorithms, the integrated approach demonstrates superior performance across key metrics. Stress detection algorithms, such as SVM, exhibit accuracy levels ranging from 70% to 96.5%, with our proposed approach achieving remarkable results. The integrated model achieves an accuracy of 96.5% and an Area under the Curve (AUC) of 0.98, surpassing existing methods in terms of accuracy, sensitivity, specificity, and AUC.

The development and productivity of maize, an important crop worldwide, may be stunted by several nutritional deficiencies. If we want to increase maize output, we need to find these problems quickly. This study suggests a thorough method for identifying nutritional deficits in maize plants by analyzing leaf photos. Our approach combines deep learning algorithms with conventional machine learning methods to analyze and extract information from these pictures. The four types of nutritional deficiencies that were examined are zinc (Zn), potassium (K), nitrogen (N), and phosphorus (P). The standard machine learning method uses Gabor, Discrete Wavelet Transform, Local Binary Pattern, and Gray-Level Co-occurrence Matrix (GLCM). Then, classification is done using algorithms like Support Vector Machine (SVM), Decision Tree, and Gradient Boosting. According to our experimental data, machine-learning algorithms successfully diagnose nutritional deficits in maize plants. The results of this study highlight the promise of machine learning algorithms for improving agricultural yields via better plant nutrition management. Farmers and agricultural specialists may greatly benefit from automated image analysis that can identify nutritional deficits in maize plants quickly and correctly. This technology has the potential to contribute to the sustainability and security of food on a worldwide scale.