Search

A new image encryption algorithm based on the triple chaotic maps is proposed to deal with the issues of inadequate security and low encryption efficiency. Coloured images consist of three linked channels used in the scheme. This method uses different keys to break the correlations between adjacent pixels in each channel. The triple chaotic maps are Lorenz, 2D‐Logistic, and Henon. First, the plain image is split into RGB channels to encrypt each channel separately. Second, the triple chaotic maps generate two groups of keys. The first group of keys performs a pixel permutation, resulting in scrambled channels used as input for the following step. Finally, the second group of keys is used to diffuse the scrambled channels independently, resulting in diffused channels, which are then merged to obtain a cipher image. The triple chaotic maps of different orders generate the cipher image with great unpredictability and security. The security is evaluated using various measures. The results demonstrated a high level of security attained by successfully encrypting coloured images. Recent encryption algorithms are compared in terms of entropy, correlation coefficients, and attack robustness. The proposed method provided outstanding security and outperformed existing image encryption algorithms. [ABSTRACT FROM AUTHOR]

The study of electric vehicles (EVs) aims to address the critical challenges of promoting widespread adoption. These challenges include EVs' high upfront costs compared to conventional vehicles, the need for more sufficient charging stations, limitations in battery technology and charging speeds, and concerns about the distance EVs can travel on a single charge. This paper is dedicated to designing an innovative strategy to handle EV charging station arrangement issues in different cities. Our research will support the development of sustainable transportation by intelligently replying to the challenges related to short ranges and long recharging times through the distribution of fast and ultra- fast charge terminals by allocating demand to charging stations while considering the cost variable of traffic congestion. A hybrid combination of Dynamic Greylag Goose Optimization (DGGO) algorithm, as well as a Long Short-Term Memory (LSTM) model, is employed in this approach to determine, in a cost-sensitive way, the location of the parking lots, factoring in the congestion for traffic as a variable. This study examines in detail the experiments on the DGGO + LSTM model performance for the purpose of finding an efficient charging station place. The results show that the DGGO + LSTM model has achieved a stunning accuracy of 0.988,836, more than the other models. This approach shapes our finding's primary purpose of proposing solutions in terms of EV charging infrastructure optimization that is fully justified to the EV's wide diffusion and mitigating of the environmental consequences. [ABSTRACT FROM AUTHOR]

The stability of smart grids is crucial for ensuring reliable and efficient power distribution in modern energy systems. This paper presents an optimized Long Short-Term Memory model for predicting smart grid stability, leveraging the Novel Guide-Waterwheel Plant Algorithm (Guide-WWPA) for enhanced performance. Traditional methods often struggle with the complexity and dynamic nature of smart grids, necessitating advanced approaches for accurate predictions. The proposed LSTM model, optimized using Guide-WWPA, addresses these challenges by effectively capturing temporal dependencies and nonlinear relationships in the data. The proposed approach involves a comprehensive preprocessing pipeline to handle data heterogeneity and noise, followed by the implementation of the LSTM model optimized through Guide-WWPA. The Guide-WWPA combines the strength of the WWPA with a novel guidance mechanism, ensuring efficient exploration and exploitation of the search space. The optimized LSTM is evaluated on a real-world smart grid dataset, demonstrating superior performance compared to traditional optimization techniques. Experimental Results indicate significant improvements in prediction accuracy and computational efficiency, highlighting the potential of the Guide-WWPA optimized LSTM for real-time smart grid stability prediction. This work contributes to the development of intelligent energy management systems, offering a robust tool for maintaining grid stability and enhancing overall energy reliability. On the other hand, statistical evaluations were carried out to prove the stability and difference of the proposed methodology. The results of the experiments demonstrate that the Guide-WWPA + LSTM strategy is superior to the other machine learning approaches. [ABSTRACT FROM AUTHOR]

As the number of individuals who drive electric vehicles increases, it is becoming increasingly important to ensure that charging infrastructure is both dependable and conveniently accessible. Methodology: In this paper, a recommendation system is proposed with the purpose of assisting users of electric vehicles in locating charging stations that are closer to them, improving the charging experience, and lowering range anxiety. The proposed method is based on restricted Boltzmann machine learning to collect and evaluate real-time data on a variety of aspects, including the availability of charging stations and historical patterns of consumption. To optimize the parameters of the restricted Boltzmann machine, a new optimization algorithm is proposed and referred to as parallel greylag goose (PGGO) algorithm. The recommendation algorithm takes into consideration a variety of user preferences. These preferences include charging speed, cost, network compatibility, amenities, and proximity to the user's present location. By addressing these preferences, the proposed approach reduces the amount of irritation experienced by users, improves charging performance, and increases customer satisfaction. Results: The findings demonstrate that the method is effective in recommending charging stations that are close to drivers of electric vehicles. On the other hand, the Wilcoxon rank-sum and Analysis of Variance tests are utilized in this work to investigate the statistical significance of the proposed parallel greylag goose optimization method and restricted Boltzmann machine model. The proposed methodology could achieve a recommendation accuracy of 99% when tested on the adopted dataset. Conclusion: Based on the achieved results, the proposed method is effective in recommending systems for the best charging stations for electric vehicles. [ABSTRACT FROM AUTHOR]

To reduce the negative effects that conventional modes of transportation have on the environment, researchers are working to increase the use of electric vehicles. The demand for environmentally friendly transportation may be hampered by obstacles such as a restricted range and extended rates of recharge. The establishment of urban charging infrastructure that includes both fast and ultra-fast terminals is essential to address this issue. Nevertheless, the powering of these terminals presents challenges because of the high energy requirements, which may influence the quality of service. Modelling the maximum hourly capacity of each station based on its geographic location is necessary to arrive at an accurate estimation of the resources required for charging infrastructure. It is vital to do an analysis of specific regional traffic patterns, such as road networks, route details, junction density, and economic zones, rather than making arbitrary conclusions about traffic patterns. When vehicle traffic is simulated using this data and other variables, it is possible to detect limits in the design of the current traffic engineering system. Initially, the binary graylag goose optimization (bGGO) algorithm is utilized for the purpose of feature selection. Subsequently, the graylag goose optimization (GGO) algorithm is utilized as a voting classifier as a decision algorithm to allocate demand to charging stations while taking into consideration the cost variable of traffic congestion. Based on the results of the analysis of variance (ANOVA), a comprehensive summary of the components that contribute to the observed variability in the dataset is provided. The results of the Wilcoxon Signed Rank Test compare the actual median accuracy values of several different algorithms, such as the voting GGO algorithm, the voting grey wolf optimization algorithm (GWO), the voting whale optimization algorithm (WOA), the voting particle swarm optimization (PSO), the voting firefly algorithm (FA), and the voting genetic algorithm (GA), to the theoretical median that would be expected that there is no difference. [ABSTRACT FROM AUTHOR]

Diabetes is a chronic disease that is usually caused when the pancreas fails to produce sufficient insulin or when the body is unable to effectively utilize the insulin produced by the pancreas. Early detection of diabetes enables the implementation of a suitable treatment method, which can lead to a healthy lifestyle. A necessity arises for an automated system capable of diagnosing diabetes using clinical and physical data in cases when the conventional approach to detecting diabetes proves to be arduous. In this paper, a new diabetes classification model based on optimized long short-term memory (LSTM) is presented and evaluated on the Pima Indians Diabetes Database (PIDD). To improve the LSTM model, a novel hybrid waterwheel plant and stochastic fractal search (WWPASFS) is proposed for optimizing its parameters. To confirm the performance superiority of the proposed WWPASFS + LSTM model, it is compared to various machine learning models and metaheuristic optimization methods. In addition, the binary WWPASFS is proposed to extract the relevant features in the PIDD dataset, with the aim of improving the accurate classification of diabetes patients. The WWPASFS + LSTM model attained the highest accuracy of 98.2% in classifying diabetes patients on the dataset in hand. The WWPASFS + LSTM model exhibited superior performance compared to the other five models, namely decision tree, K-nearest neighbors, neural networks, random forest, and support vector machines. On the other hand, the statistical analysis of the proposed approach is studied and the results prove its difference and significance. [ABSTRACT FROM AUTHOR]

The networks of wireless sensors provide the ground for a range of applications, including environmental monitoring and industrial operations. Ensuring the networks can overcome obstacles like power and communication reliability and sensor coverage is the crux of network optimization. Network infrastructure planning should be focused on increasing performance, and it should be affected by the detailed data about node distribution. This work recommends the creation of each sensor's specs and radius of influence based on a particular geographical location, which will contribute to better network planning and design. By using the ARIMA model for time series forecasting and the Al-Biruni Earth Radius algorithm for optimization, our approach bridges the gap between successive terrains while seeking the equilibrium between exploration and exploitation. Through implementing adaptive protocols according to varying environments and sensor constraints, our study aspires to improve overall network operation. We compare the Al-Biruni Earth Radius algorithm along with Gray Wolf Optimization, Particle Swarm Optimization, Genetic Algorithms, and Whale Optimization about performance on real-world problems. Being the most efficient in the optimization process, Biruni displays the lowest error rate at 0.00032. The two other statistical techniques, like ANOVA, are also useful in discovering the factors influencing the nature of sensor data and network-specific problems. Due to the multi-faceted support the comprehensive approach promotes, there is a chance to understand the dynamics that affect the optimization outcomes better so decisions about network design can be made. Through delivering better performance and reliability for various in-situ applications, this research leads to a fusion of time series forecasters and a customized optimizer algorithm. [ABSTRACT FROM AUTHOR]

In the literature, it is well known that there is a bidirectional causality between economic growth and energy consumption. This is why it is crucial to forecast energy consumption. In this study, four deep learning models, i.e., Long Short-Term Memory (LSTM), stacked LSTM, bidirectional LSTM, and Gated Recurrent Unit (GRU), were used to forecast energy consumption in Brazil, Canada, and France. After a training test period, the performance evaluation criterion, i.e., R2, mean square error, root mean square error, mean absolute error, and mean absolute percentage error, was performed for the performance measure. It showed that GRU is the best model for Canada and France, while LSTM is the best model for Brazil. Therefore, the energy consumption prediction was made for the 12 months of the year 2017 using LSTM for Brazil and GRU for Canada and France. Based on the selected model, it was projected that the energy consumption in Brazil was 38 597.14–38 092.88, 63 900–4 800 000 GWh in Canada, and 50 999.72–32 747.01 GWh in France in 2017. The projected consumption in Canada was very high due to the country's higher industrialization. The results obtained in this study confirmed that the nature of energy production will impact the complexity of the deep learning model. [ABSTRACT FROM AUTHOR]

Energy consumption in buildings is gradually increasing and accounts for around forty percent of the total energy consumption. Forecasting the heating and cooling loads of a building during the initial phase of the design process in order to identify optimal solutions among various designs is of utmost importance. This is also true during the operation phase of the structure after it has been completed in order to ensure that energy efficiency is maintained. The aim of this paper is to create and develop a Multilayer Perceptron Regressor (MLPRegressor) model for the purpose of forecasting the heating and cooling loads of a building. The proposed model is based on automated hyperparameter optimization using Waterwheel Plant Algorithm The model was based on a dataset that described the energy performance of the structure. There are a number of important characteristics that are considered to be input variables. These include relative compactness, roof area, overall height, surface area, glazing area, wall area, glazing area distribution of a structure, and orientation. On the other hand, the variables that are considered to be output variables are the heating and cooling loads of the building. A total of 768 residential buildings were included in the dataset that was utilized for training purposes. Following the training and regression of the model, the most significant parameters that influence heating load and cooling load have been identified, and the WWPA- MLPRegressor performed well in terms of different metrices variables and fitted time. [ABSTRACT FROM AUTHOR]

The use of privacy-enhanced facial recognition has increased in response to growing concerns about data security and privacy in the digital age. This trend is spurred by rising demand for face recognition technology in a variety of industries, including access control, law enforcement, surveillance, and internet communication. However, the growing usage of face recognition technology has created serious concerns about data monitoring and user privacy preferences, especially in context-aware systems. In response to these problems, this study provides a novel framework that integrates sophisticated approaches such as Generative Adversarial Networks (GANs), Blockchain, and distributed computing to solve privacy concerns while maintaining exact face recognition. The framework's painstaking design and execution strive to strike a compromise between precise face recognition and protecting personal data integrity in an increasingly interconnected environment. Using cutting-edge tools like Dlib for face analysis,Ray Cluster for distributed computing, and Blockchain for decentralized identity verification, the proposed system provides scalable and secure facial analysis while protecting user privacy. The study's contributions include the creation of a sustainable and scalable solution for privacy-aware face recognition, the implementation of flexible privacy computing approaches based on Blockchain networks, and the demonstration of higher performance over previous methods. Specifically, the proposed StyleGAN model has an outstanding accuracy rate of 93.84% while processing high-resolution images from the CelebA-HQ dataset, beating other evaluated models such as Progressive GAN 90.27%, CycleGAN 89.80%, and MGAN 80.80%. With improvements in accuracy, speed, and privacy protection, the framework has great promise for practical use in a variety of fields that need face recognition technology. This study paves the way for future research in privacy-enhanced face recognition systems, emphasizing the significance of using cutting-edge technology to meet rising privacy issues in digital identity. [ABSTRACT FROM AUTHOR]

Nanoemulsions are gaining interest in a variety of products as a means of integrating easily degradable bioactive compounds, preserving them from oxidation, and increasing their bioavailability. However, preparing stable emulsion compositions with the desired characteristics is a difficult task. The aim of this study was to encapsulate the Tinospora cordifolia aqueous extract (TCAE) into a water in oil (W/O) nanoemulsion and identify its critical process and formulation variables, like oil (27–29.4 mL), the surfactant concentration (0.6–3 mL), and sonication amplitude (40% to 100%), using response surface methodology (RSM). The responses of this formulation were studied with an analysis of the particle size (PS), free fatty acids (FFAs), and encapsulation efficiency (EE). In between, we have studied a fishbone diagram that was used to measure risk and preliminary research. The optimized condition for the formation of a stable nanoemulsion using quality by design was surfactant (2.43 mL), oil concentration (27.61 mL), and sonication amplitude (88.6%), providing a PS of 171.62 nm, FFA content of 0.86 meq/kg oil and viscosity of 0.597 Pa.s for the blank sample compared to the enriched TCAE nanoemulsion with a PS of 243.60 nm, FFA content of 0.27 meq/kg oil and viscosity of 0.22 Pa.s. The EE increases with increasing concentrations of TCAE, from 56.88% to 85.45%. The RSM response demonstrated that both composition variables had a considerable impact on the properties of the W/O nanoemulsion. Furthermore, after the storage time, the enriched TCAE nanoemulsion showed better stability over the blank nanoemulsion, specially the FFAs, and the blank increased from 0.142 to 1.22 meq/kg oil, while TCAE showed 0.266 to 0.82 meq/kg. [ABSTRACT FROM AUTHOR]

A new metaheuristic called the Walrus Optimizer (WO) is inspired by the ways in which walruses move, roost, feed, spawn, gather, and flee in response to important cues (safety and danger signals). In this work, the WO was used to address the economic load dispatch (ELD) issue, which is one of the essential parts of a power system. One type of ELD was designed to reduce fuel consumption expenses. A variety of methodologies were used to compare the WO's performance in order to determine its reliability. These methods included rime-ice algorithm (RIME), moth search algorithm (MSA), the snow ablation algorithm (SAO), and chimp optimization algorithm (ChOA) for the identical case study. We employed six scenarios: Six generators operating at two loads of 700 and 1000 MW each were employed in the first two cases for the ELD problem. For the ELD problem, the second two scenarios involved ten generators operating at two loads of 2000 MW and 1000 MW. Twenty generators operating at a 3000 MW load were the five cases for the ELD issue. Thirty generators operating at a 5000 MW load were the six cases for the ELD issue. The power mismatch factor was the main cause of ELD problems. The ideal value of this component should be close to zero. Using the WO approach, the ideal power mismatch values of 4.1922E-13 and 4.5119E-13 were found for six generator units at demand loads of 700 MW and 1000 MW, respectively. Using metrics for the minimum, mean, maximum, and standard deviation of fitness function, the procedures were evaluated over thirty separate runs. The WO outperformed all other algorithms, as seen by the results generated for the six ELD case studies. [ABSTRACT FROM AUTHOR]

Lead-based deep brain stimulation (DBS) electrodes have been employed to treat Parkinson's disease (PD), but their limitations have led to the development of lead-free piezoelectric nanoparticle-based DBS (LF-PND-DBS). This novel approach utilizes non-invasive biocompatible piezoelectric nanoparticles to generate electrical stimulation, offering a promising alternative to traditional DBS. In this study, an innovative machine learning (ML)-optimized LF-PND-DBS system for diagnosing and evaluating PD is proposed. By leveraging ML algorithms, the optimized design of LF-PND electrodes and stimulation parameters is derived, ensuring precise and personalized treatment delivery. The ML-optimized LF-PND-DBS system was evaluated in a cohort of PD patients, demonstrating an exceptional diagnostic accuracy with a sensitivity of 99.1% and a specificity of 98.2%. It effectively assessed PD severity and response to DBS treatment, providing valuable guidance for treatment monitoring. The findings highlight the immense potential of the ML-optimized LF-PND-DBS system as a transformative tool for PD diagnosis and evaluation. This novel approach has the potential to enhance DBS efficacy, safety, and personalization, paving the way for improved patient outcomes and quality of life. [ABSTRACT FROM AUTHOR]

The rising popularity of electric vehicles (EVs) can be attributed to their positive impact on the environment and their ability to lower operational expenses. Nevertheless, the task of determining the most suitable EV types for a specific site continues to pose difficulties, mostly due to the wide range of consumer preferences and the inherent limits of EVs. This study introduces a new voting classifier model that incorporates the Al-Biruni earth radius optimization algorithm, which is derived from the stochastic fractal search. The model aims to predict the optimal EVtype for a given location by considering factors such as user preferences, availability of charging infrastructure, and distance to the destination. The proposed classification methodology entails the utilization of ensemble learning, which can be subdivided into two distinct stages: pre-classification and classification. During the initial stage of classification, the process of data preprocessing involves converting unprocessed data into a refined, systematic, and well-arranged format that is appropriate for subsequent analysis or modeling. During the classification phase, a majority vote ensemble learning method is utilized to categorize unlabeled data properly and efficiently. This method consists of three independent classifiers. The efficacy and efficiency of the suggested method are showcased through simulation experiments. The results indicate that the collaborative classification method performs very well and consistently in classifying EV populations. In comparison to similar classification approaches, the suggested method demonstrates improved performance in terms of assessment metrics such as accuracy, sensitivity, specificity, and F-score. The improvements observed in these metrics are 91.22%, 94.34%, 89.5%, and 88.5%, respectively. These results highlight the overall effectiveness of the proposed method. Hence, the suggested approach is seen more favorable for implementing the voting classifier in the context of the EV population across different geographical areas. [ABSTRACT FROM AUTHOR]

Tomato leaf diseases significantly impact crop production, necessitating early detection for sustainable farming. Deep Learning (DL) has recently shown excellent results in identifying and classifying tomato leaf diseases. However, current DL methods often require substantial computational resources, hindering their application on resource-constrained devices.We propose theDeepTomatoDetectionNetwork (DTomatoDNet), a lightweightDLbased framework comprising 19 learnable layers for efficient tomato leaf disease classification to overcome this. The Convn kernels used in the proposed (DTomatoDNet) framework is 1×1, which reduces the number of parameters and helps inmore detailed and descriptive feature extraction for classification. The proposedDTomatoDNetmodel is trained from scratch to determine the classification success rate. 10,000 tomato leaf images (1000 images per class) from the publicly accessible dataset, covering one healthy category and nine disease categories, are utilized in training the proposed DTomatoDNet approach. More specifically, we classified tomato leaf images into Target Spot (TS), Early Blight (EB), Late Blight (LB), Bacterial Spot (BS), LeafMold (LM), Tomato Yellow Leaf Curl Virus (YLCV), Septoria Leaf Spot (SLS), Spider Mites (SM), Tomato Mosaic Virus (MV), and Tomato Healthy (H). The proposedDTomatoDNet approach obtains a classification accuracy of 99.34%, demonstrating excellent accuracy in differentiating between tomato diseases.Themodel could be used onmobile platforms because it is lightweight and designed with fewer layers. Tomato farmers can utilize the proposed DTomatoDNetmethodology to detect disease more quickly and easily once it has been integrated into mobile platforms by developing a mobile application. [ABSTRACT FROM AUTHOR]

With the capacities of self-learning, acquainted capacities, high-speed looking for ideal arrangements, solid nonlinear fitting, and mapping self-assertively complex nonlinear relations, neural systems have made incredible advances and accomplished broad application over the final half-century. As one of the foremost conspicuous methods for fake insights, neural systems are growing toward high computational speed andmoo control utilization. Due to the inborn impediments of electronic gadgets, it may be troublesome for electronic-implemented neural systems to make the strides these two exhibitions encourage. Optical neural systems can combine optoelectronic procedures and neural organizationmodels to provide ways to break the bottleneck. This paper outlines optical neural networks of feedforward repetitive and spiking models to give a clearer picture of history, wildernesses, and future optical neural systems. The framework demonstrates neural systems in optic communication with the serial and parallel setup. The graphene-based laser structure for fiber optic communication is discussed. The comparison of different balance plans for photonic neural systems is made within the setting of hereditary calculation and molecule swarm optimization. In expansion, the execution comparison of routine photonic neural, time-domain with and without extending commotion is additionally expounded. The challenges and future patterns of optical neural systems on the growing scale and applications of in situ preparing nonlinear computing will hence be uncovered. [ABSTRACT FROM AUTHOR]

Face mask detection has several applications, including real-time surveillance, biometrics, etc. Identifying face masks is also helpful for crowd control and ensuring people wear them publicly. With monitoring personnel, it is impossible to ensure that people wear face masks; automated systems are a much superior option for face mask detection and monitoring. This paper introduces a simple and efficient approach for masked face detection. The architecture of the proposed approach is very straightforward; it combines deep learning and local binary patterns to extract features and classify them as masked or unmasked. The proposed system requires hardware with minimal power consumption compared to state-of-the-art deep learning algorithms. Our proposed system maintains two steps. At first, this work extracted the local features of an image by using a local binary pattern descriptor, and then we used deep learning to extract global features. The proposed approach has achieved excellent accuracy and high performance. The performance of the proposed method was tested on three benchmark datasets: the realworld masked faces dataset (RMFD), the simulated masked faces dataset (SMFD), and labeled faces in the wild (LFW). Performancemetrics for the proposed technique weremeasured in terms of accuracy, precision, recall, and F1-score. Results indicated the efficiency of the proposed technique, providing accuracies of 99.86%, 99.98%, and 100% for RMFD, SMFD, and LFW, respectively. Moreover, the proposed method outperformed state-of-the-art deep learningmethods in the recent bibliography for the same problem under study and on the same evaluation datasets. [ABSTRACT FROM AUTHOR]

The paper focuses on the hepatitis C virus (HCV) infection in Egypt, which has one of the highest rates of HCV in the world. The high prevalence is linked to several factors, including the use of injection drugs, poor sterilization practices in medical facilities, and low public awareness. This paper introduces a hyOPTGB model, which employs an optimized gradient boosting (GB) classifier to predict HCV disease in Egypt. The model's accuracy is enhanced by optimizing hyperparameters with the OPTUNA framework. Min-Max normalization is used as a preprocessing step for scaling the dataset values and using the forward selection (FS) wrapped method to identify essential features. The dataset used in the study contains 1385 instances and 29 features and is available at the UCI machine learning repository. The authors compare the performance of five machine learning models, including decision tree (DT), support vector machine (SVM), dummy classifier (DC), ridge classifier (RC), and bagging classifier (BC), with the hyOPTGB model. The system's efficacy is assessed using various metrics, including accuracy, recall, precision, and F1-score. The hyOPTGB model outperformed the other machine learning models, achieving a 95.3% accuracy rate. The authors also compared the hyOPTGB model against other models proposed by authors who used the same dataset. [ABSTRACT FROM AUTHOR]

Orthogonal generalized Laguerre moments of fractional orders (FrGLMs) are signal and image descriptors. The utilization of the FrGLMs in the analysis of big-size signals encounters three challenges. First, calculating the high-order moments is a time-consuming process. Second, accumulating numerical errors leads to numerical instability and degrades the reconstructed signals' quality. Third, the QR decomposition technique is needed to preserve the orthogonality of the higher-order moments. In this paper, the authors derived a new recurrence formula for calculating the FrGLMs, significantly reducing the computational CPU times. We used the Schwarz–Rutishauser algorithm as an alternative to the QR decomposition technique. The proposed method for computing FrGLMs for big-size signals is accurate, simple, and fast. The proposed algorithm has been tested using the MIT-BIH arrhythmia benchmark dataset. The results show the proposed method's superiority over existing methods in terms of processing time and reconstruction capability. Concerning the reconstructed capability, it has achieved superiority with average values of 25.3233 and 15.6507 with the two metrics PSNR and MSE, respectively. Concerning the elapsed reconstruction time, it also achieved high superiority with an efficiency gain of 0.8. The proposed method is suitable for utilization in the Internet of Healthcare Things. [ABSTRACT FROM AUTHOR]

Image authentication techniques have recently received a lot of attention for protecting images against unauthorized access. Due to the wide use of the Internet nowadays, the need to ensure data integrity and authentication increases. Many techniques, such as watermarking and encryption, are used for securing images transmitted via the Internet. The majority of watermarking systems are PC-based, but they are not very portable. Hardwarebased watermarkingmethods need to be developed to accommodate real-time applications and provide portability. This paper presents hybrid data security techniques using a zero watermarkingmethod to provide copyright protection for the transmitted color images using multi-channel orthogonal Legendre Fourier moments of fractional orders (MFrLFMs) and the advanced encryption standard (AES) algorithm on a low-cost Raspberry Pi. In order to increase embedding robustness, the watermark picture is scrambled using the Arnold method. Zero watermarking is implemented on the Raspberry Pi to produce a real-time ownership verification key. Before sending the ownership verification key and the original image to the monitoring station, we can encrypt the transmitted data with AES for additional security and hide any viewable information. The receiver next verifies the received image's integrity to confirmits authenticity and that it has not been tampered with. We assessed the suggested algorithm's resistance tomany attacks. The suggested algorithm provides a reasonable degree of robustness while still being perceptible. The proposed method provides improved bit error rate (BER) and normalized correlation (NC) values compared to previous zero watermarking approaches. AES performance analysis is performed to demonstrate its effectiveness. Using a 256 × 256 image size, it takes only 2 s to apply the zero-watermark algorithm on the Raspberry Pi. [ABSTRACT FROM AUTHOR]

The COVID-19 epidemic poses a worldwide threat that transcends provincial, philosophical, spiritual, radical, social, and educational borders. By using a connected network, a healthcare system with the Internet of Things (IoT) functionality can effectively monitor COVID-19 cases. IoT helps a COVID-19 patient recognize symptoms and receive better therapy more quickly. A critical component in measuring, evaluating, and diagnosing the risk of infection is artificial intelligence (AI). It can be used to anticipate cases and forecast the alternate incidences number, retrieved instances, and injuries. In the context of COVID-19, IoT technologies are employed in specific patient monitoring and diagnosing processes to reduce COVID-19 exposure to others. This work uses an Indian dataset to create an enhanced convolutional neural network with a gated recurrent unit (CNN-GRU) model for COVID-19 death prediction via IoT. The data were also subjected to data normalization and data imputation. The 4692 cases and eight characteristics in the dataset were utilized in this research. The performance of the CNN-GRU model for COVID-19 death prediction was assessed using five evaluation metrics, including median absolute error (MedAE), mean absolute error (MAE), root mean squared error (RMSE), mean square error (MSE), and coefficient of determination (R2). ANOVA and Wilcoxon signed-rank tests were used to determine the statistical significance of the presented model. The experimental findings showed that the CNN-GRU model outperformed other models regarding COVID-19 death prediction. [ABSTRACT FROM AUTHOR]

This study introduces ETLBOCBL-CNN, an automated approach for optimizing convolutional neural network (CNN) architectures to address classification tasks of varying complexities. ETLBOCBL-CNN employs an effective encoding scheme to optimize network and learning hyperparameters, enabling the discovery of innovative CNN structures. To enhance the search process, it incorporates a competency-based learning concept inspired by mixed-ability classrooms during the teacher phase. This categorizes learners into competency-based groups, guiding each learner's search process by utilizing the knowledge of the predominant peers, the teacher solution, and the population mean. This approach fosters diversity within the population and promotes the discovery of innovative network architectures. During the learner phase, ETLBOCBL-CNN integrates a stochastic peer interaction scheme that encourages collaborative learning among learners, enhancing the optimization of CNN architectures. To preserve valuable network information and promote long-term population quality improvement, ETLBOCBL-CNN introduces a tri-criterion selection scheme that considers fitness, diversity, and learners' improvement rates. The performance of ETLBOCBL-CNN is evaluated on nine different image datasets and compared to state-of-the-art methods. Notably, ELTLBOCBL-CNN achieves outstanding accuracies on various datasets, including MNIST (99.72%), MNIST-RD (96.67%), MNIST-RB (98.28%), MNIST-BI (97.22%), MNST-RD + BI (83.45%), Rectangles (99.99%), Rectangles-I (97.41%), Convex (98.35%), and MNIST-Fashion (93.70%). These results highlight the remarkable classification accuracy of ETLBOCBL-CNN, underscoring its potential for advancing smart device infrastructure development. [ABSTRACT FROM AUTHOR]

One of the significant health issues affecting women that impacts their fertility and results in serious health concerns is Polycystic ovarian syndrome (PCOS). Consequently, timely screening of polycystic ovarian syndrome can help in the process of recovery. Finding a method to aid doctors in this procedure was crucial due to the difficulties in detecting this condition. This research aimed to determine whether it is possible to optimize the detection of PCOS utilizing Deep Learning algorithms and methodologies. Additionally, feature selection methods that produce the most important subset of features can speed up calculation and enhance the effectiveness of classifiers. In this research, the tri-stage wrapper method is used because it reduces the computation time. The proposed study for the Automatic diagnosis of PCOS contains preprocessing, data normalization, feature selection, and classification. A dataset with 39 characteristics, including metabolism, neuroimaging, hormones, and biochemical information for 541 subjects, was employed in this scenario. To start, this research pre-processed the information. Next for feature selection, a tri-stage wrapper method such as Mutual Information, ReliefF, Chi-Square, and Xvariance is used. Then, various classification methods are tested and trained. Deep learning techniques including convolutional neural network (CNN), multi-layer perceptron (MLP), Recurrent neural network (RNN), and Bi long short-term memory (Bi-LSTM) are utilized for categorization. The experimental finding demonstrates that with effective feature extraction process using tri stage wrapper method + CNN delivers the highest precision (97%), high accuracy (98.67%), and recall (89%) when compared with other machine learning algorithms. [ABSTRACT FROM AUTHOR]

Test case selection is to minimize the time and effort spent on software testing in real-time practice. During software testing, software firms need techniques to finish the testing in a stipulated time while uncompromising on quality. The motto is to select a subset of test cases rather than take up all available test cases to uncover most bugs. Our proposed model in the research study effort is termed SCARF-RT, which stands for Similarity coefficient (SC), Creating Acronyms, Regression test (RT), and Fuzzy set (FS) with Dataset (DS). Clustering of test cases using ranking and also based on similarity coefficients is to be implemented. This research considered eleven different features for clustering the test cases. Two techniques have been used. Firstly, each cluster will, to a certain extent, encompass a collection of distinct traits. Depending on the coverage of the feature, a cluster of test cases might be chosen. The ranking approach was used to create these groupings. The second methodology finds similarity among test cases based on eleven features. Then, the maxmin composition is used to find fuzzy equivalences upon which clusters are formed. Most similar test cases are clustered. Test cases of every cluster are selected as a test suite. The outcomes of this research show that the selected test cases based on the proposed approaches are better than existing methodologies in selecting test cases with less duration and at the same time not compromising on quality. Both fuzzy rank-based clustering and similarity coefficient-based clustering test case selection approaches have been developed and implemented. With the help of these methods, testers may quickly choose test cases based on the suggested characteristics and complete regression testing more quickly. [ABSTRACT FROM AUTHOR]

Accurate cyclone intensity prediction is crucial for smart cities to effectively prepare and mitigate the potential devastation caused by these extreme weather events. Traditional meteorological models often face challenges in accurately forecasting cyclone intensity due to cyclonic systems' complex and dynamic nature. Predicting the intensity of cyclones is a challenging task in meteorological research, as it requires expertise in extracting spatio-temporal features. To address this challenge, a new technique, called linear support vector regressive gradient descent Jaccardized deep multilayer perceptive classifier (LEGEMP), has been proposed to improve the accuracy of cyclone intensity prediction. This technique utilizes a dataset that contains various attributes. It employs the Herfindahl correlative linear support vector regression feature selection to identify the most important characteristics for enhancing cyclone intensity forecasting accuracy. The selected features are then used in conjunction with the Nesterov gradient descent jeopardized deep multilayer perceptive classifier to predict the intensity classes of cyclones, including depression, deep depression, cyclone, severe cyclone, very severe cyclone, and extremely severe cyclone. Experimental results have demonstrated that LEGEMP outperforms conventional methods in terms of cyclone intensity prediction accuracy, requiring minimum time, error rate, and memory consumption. By leveraging advanced techniques and feature selection, LEGEMP provides more reliable and precise predictions for cyclone intensity, enabling better preparedness and response strategies to mitigate the impact of these destructive storms. The LEGEMP technique offers an improved approach to cyclone intensity prediction, leveraging advanced classifiers and feature selection methods to enhance accuracy and reduce error rates. We demonstrate the effectiveness of our approach through rigorous evaluation and comparison with conventional prediction methods, showcasing significant improvements in prediction accuracy. Integrating our enhanced prediction model into smart city disaster management systems can substantially enhance preparedness and response strategies, ultimately contributing to the safety and resilience of communities in cyclone-prone regions. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have excelled in artificial intelligence, particularly in image-related tasks such as classification and object recognition. However, manually designing CNN architectures demands significant domain expertise and involves time-consuming trial-and-error processes, along with substantial computational resources. To overcome this challenge, an automated network design method known as Modified Teaching-Learning-Based Optimization with Refined Knowledge Sharing (MTLBORKS-CNN) is introduced. It autonomously searches for optimal CNN architectures, achieving high classification performance on specific datasets without human intervention. MTLBORKS-CNN incorporates four key features. It employs an effective encoding scheme for various network hyperparameters, facilitating the search for innovative and valid network architectures. During the modified teacher phase, it leverages a social learning concept to calculate unique exemplars that effectively guide learners while preserving diversity. In the modified learner phase, self-learning and adaptive peer learning are incorporated to enhance knowledge acquisition of learners during CNN architecture optimization. Finally, MTLBORKS-CNN employs a dual-criterion selection scheme, considering both fitness and diversity, to determine the survival of learners in subsequent generations. MTLBORKS-CNN is rigorously evaluated across nine image datasets and compared with state-of-the-art methods. The results consistently demonstrate MTLBORKS-CNN's superiority in terms of classification accuracy and network complexity, suggesting its potential for infrastructural development of smart devices. [ABSTRACT FROM AUTHOR]

The osprey optimization algorithm (OOA) is a new metaheuristic motivated by the strategy of hunting fish in seas. In this study, the OOA is applied to solve one of the main items in a power system called economic load dispatch (ELD). The ELD has two types. The first type takes into consideration the minimization of the cost of fuel consumption, this type is called ELD. The second type takes into consideration the cost of fuel consumption and the cost of emission, this type is called combined emission and economic dispatch (CEED). The performance of the OOA is compared against several techniques to evaluate its reliability. These methods include elephant herding optimization (EHO), the rime-ice algorithm (RIME), the tunicate swarm algorithm (TSA), and the slime mould algorithm (SMA) for the same case study. Also, the OOA is compared with other techniques in the literature, such as an artificial bee colony (ABO), the sine cosine algorithm (SCA), the moth search algorithm (MSA), the chimp optimization algorithm (ChOA), and monarch butterfly optimization (MBO). Power mismatch is the main item used in the evaluation of the OOA with all of these methods. There are six cases used in this work: 6 units for the ELD problem at three different loads, and 6 units for the CEED problem at three different loads. Evaluation of the techniques was performed for 30 various runs based on measuring the standard deviation, minimum fitness function, and maximum mean values. The superiority of the OOA is achieved according to the obtained results for the ELD and CEED compared to all competitor algorithms. [ABSTRACT FROM AUTHOR]

Brain neoplasms are recognized with a biopsy, which is not commonly done before decisive brain surgery. By using Convolutional Neural Networks (CNNs) and textural features, the process of diagnosing brain tumors by radiologists would be a noninvasive procedure. This paper proposes a features fusion model that can distinguish between no tumor and brain tumor types via a novel deep learning structure. The proposed model extracts Gray Level Co-occurrence Matrix (GLCM) textural features from MRI brain tumor images. Moreover, a deep neural network (DNN) model has been proposed to select the most salient features from the GLCM. Moreover, it manipulates the extraction of the additional high levels of salient features from a proposed CNN model. Finally, a fusion process has been utilized between these two types of features to form the input layer of additional proposed DNN model which is responsible for the recognition process. Two common datasets have been applied and tested, Br35H and FigShare datasets. The first dataset contains binary labels, while the second one splits the brain tumor into four classes; glioma, meningioma, pituitary, and no cancer. Moreover, several performance metrics have been evaluated from both datasets, including, accuracy, sensitivity, specificity, F-score, and training time. Experimental results show that the proposed methodology has achieved superior performance compared with the current state of art studies. The proposed system has achieved about 98.22% accuracy value in the case of the Br35H dataset however, an accuracy of 98.01% has been achieved in the case of the FigShare dataset. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) systems are increasingly used in corporate security measures to predict the status of assets and suggest appropriate procedures. These programs are also designed to reduce repair time. One way to create an efficient system is to integrate physical repair agents with a computerized management system to develop an intelligent system. To address this, there is a need for a new technique to assist operators in interacting with a predictive system using natural language. The system also uses double neural network convolutional models to analyze device data. For fault prioritization, a technique utilizing fuzzy logic is presented. This strategy ranks the flaws based on the harm or expense they produce. However, the method's success relies on ongoing improvement in spoken language comprehension through language modification and query processing. To carry out this technique, a conversation-driven design is necessary. This type of learning relies on actual experiences with the assistants to provide efficient learning data for language and interaction models. These models can be trained to have more natural conversations. To improve accuracy, academics should construct and maintain publicly usable training sets to update word vectors. We proposed the model dataset (DS) with the Adam (AD) optimizer, Ridge Regression (RR) and Feature Mapping (FP). Our proposed algorithm has been coined with an appropriate acronym DSADRRFP. The same proposed approach aims to leverage each component's benefits to enhance the predictive model's overall performance and precision. This ensures the model is up-to-date and accurate. In conclusion, an AI system integrated with physical repair agents is a useful tool in corporate security measures. However, it needs to be refined to extract data from the operating system and to interact with users in a natural language. The system also needs to be constantly updated to improve accuracy. [ABSTRACT FROM AUTHOR]

Parkinson's disease (PD) affects a large proportion of elderly people. Symptoms include tremors, slow movement, rigid muscles, and trouble speaking. With the aging of the developed world's population, this number is expected to rise. The early detection of PD and avoiding its severe consequences require a precise and efficient system. Our goal is to create an accurate AI model that can identify PD using human voices. We developed a transformer-based method for detecting PD by retrieving dysphonia measures from a subject's voice recording. It is uncommon to use a neural network (NN)-based solution for tabular vocal characteristics, but it has several advantages over a tree-based approach, including compatibility with continuous learning and the network's potential to be linked with an image/voice encoder for a more accurate multi modal solution, shifting SOTA approach from tree-based to a neural network (NN) is crucial for advancing research in multimodal solutions. Our method outperforms the state of the art (SOTA), namely Gradient-Boosted Decision Trees (GBDTs), by at least 1% AUC, and the precision and recall scores are also improved. We additionally offered an XgBoost-based feature-selection method and a fully connected NN layer technique for including continuous dysphonia measures, in addition to the solution network. We also discussed numerous important discoveries relating to our suggested solution and deep learning (DL) and its application to dysphonia measures, such as how a transformer-based network is more resilient to increased depth compared to a simple MLP network. The performance of the proposed approach and conventional machine learning techniques such as MLP, SVM, and Random Forest (RF) have also been compared. A detailed performance comparison matrix has been added to this article, along with the proposed solution's space and time complexity. [ABSTRACT FROM AUTHOR]

The accurate prediction of energy consumption has effective role in decision making and risk management for individuals and governments. Meanwhile, the accurate prediction can be realized using the recent advances in machine learning and predictive models. This research proposes a novel approach for energy consumption forecasting based on a new optimization algorithm and a new forecasting model consisting of a set of long shortterm memory (LSTM) units. The proposed optimization algorithm is used to optimize the parameters of the LSTM-based model to boost its forecasting accuracy. This optimization algorithm is based on the recently emerged dipper-throated optimization (DTO) and stochastic fractal search (SFS) algorithm and is referred to as dynamic DTOSFS. To prove the effectiveness and superiority of the proposed approach, five standard benchmark algorithms, namely, stochastic fractal search (SFS), dipper throated optimization (DTO), whale optimization algorithm (WOA), particle swarm optimization (PSO), and grey wolf optimization (GWO), are used to optimize the parameters of the LSTM-based model, and the results are compared with that of the proposed approach. Experimental results show that the proposed DDTOSFS+LSTM can accurately forecast the energy consumption with root mean square error RMSE of 0.00013, which is the best among the recorded results of the other methods. In addition, statistical experiments are conducted to prove the statistical difference of the proposed model. The results of these tests confirmed the expected outcomes. [ABSTRACT FROM AUTHOR]