The authoritative source for research information and outputs is Elsevier Scopus, with researcher profile verification from the Deanship of Scientific Research.
المخرجات
Research outputs
Articles, reviews, conference papers and book chapters, indexed by DOI and linked to University researchers.
Process mining, a burgeoning discipline within data science, demonstrates a significant contribution to the software development lifecycle of diverse real-time consumer-centric projects. This paper underscores the prominence of integrating predictive business process monitoring into organizational process models, as it can substantially impact profits and efficiency in any possible business domain along with improving services to consumers. The paper proposes a novel deep learning-based business process prediction model consisting of multiple layers with fine-tuning hyperparameters. The proposed model leverages input embeddings to represent each of the activities, and based on the training of the proposed model, the accuracy of the next activity is calculated. To assess the efficacy of the proposed model, it has been compared with the existing benchmark models. Our proposed model has shown a significant gain over the existing approaches. The results show that the proposed model outperforms these approaches by achieving an accuracy of 76% on the consumer helpdesk dataset along with an accuracy of 78% on the benchmark BPI dataset.
process miningpredictive monitoringnext activity predictiondeep learning
Abstract In the realm of Drug‐Target Interaction (DTI) prediction, this research investigates and contrasts the efficacy of diverse evolutionary algorithms in fine‐tuning a sophisticated hybrid deep learning model. Recognizing the critical role of DTI in drug discovery and repositioning, we tackle the challenges of binary classification by reframing the problem as a regression task. Our focus lies on the Convolution Self‐Attention Network with Attention‐based bidirectional Long Short‐Term Memory Network (CSAN‐BiLSTM‐Att), a hybrid model combining convolutional neural network (CNN) blocks, self‐attention mechanisms, and bidirectional LSTM layers. To optimize this complex model, we employ Differential Evolution (DE), Particle Swarm Optimization (PSO), Memetic Particle Swarm Optimization Algorithm (MPSOA), Fire Hawk Optimization (FHO), and Artificial Hummingbird Algorithm (AHA). Through thorough comparative analysis, we evaluate the performance of these evolutionary algorithms in enhancing the CSAN‐BiLSTM‐Att model's effectiveness. By examining the strengths and weaknesses of each algorithm, our study aims to provide valuable insights into DTI prediction, identifying the most effective evolutionary algorithm for hyperparameter tuning in advanced deep learning models. Notably, Fire‐hawk optimization (FHO) emerges as particularly promising, achieving the highest Concordance Index (C‐index) as 0.974 for KIBA datasets and 0.894 for DAVIS datasets and demonstrating exceptional accuracy in ranking continuous predictions across both the datasets.
drug-target interaction predictionevolutionary algorithmshyperparameter optimizationhybrid deep learning model
Abstract Autonomous intelligent systems are artificial intelligence (AI) tools that act autonomously without direct human supervision. Cloud computing (CC) and Internet of Things (IoT) technologies find it challenging to deploy sufficient security defences because of the different structures, storage, and limited computing capabilities that make them more vulnerable to attacks. Security threats against IoT structures, devices, and applications are increasing with the demand for IoT technology. The training data available to AI models may be limited, which could impact their performance and generalizability. Adopting AI solutions in real‐world situations may be impeded by compatibility concerns and the requirement for flawless integration. Malware classification errors can occur due to a lack of contextual knowledge, particularly in cases where benign files behave identically to malicious. Various studies were carried out on detecting IoT malware to evade the menaces posed by malicious code. However, prevailing techniques of IoT malware classification supported particular platforms or demanded complicated methods for attaining higher accuracy. This study introduces an enhanced slime mould optimization with a convolutional BLSTM autoencoder‐based malware classification (ESMO‐CBLSTMAE) system in the IoT cloud platform. The projected ESMO‐CBLSTMAE system focuses on detecting and classifying malware in the IoT cloud platform. To achieve that, the ESMO‐CBLSTMAE algorithm employs a min–max normalization technique for scaling the input dataset. The ESMO‐CBLSTMAE method uses a convolutional bidirectional long short‐term memory autoencoder (CBLSTM‐AE) model for the malware detection process. Lastly, the ESMO method is executed for the optimum hyperparameter tuning of the CBLSTM‐AE technique, which boosts the malware classification results. The experimental analysis of the ESMO‐CBLSTMAE method is tested against a benchmark database, and the outcomes portray the greater efficacy of the ESMO‐CBLSTMAE approach over other existing techniques. The proposed malware classification model achieved an accuracy of 98.57 and F Score of 80.77 and outperformed the existing models.
The progress of autonomous transport systems (ATS) involves efficient multimedia communication for real-time data tradeoffs and environmental issues. Deep learning (DL) powered visual place recognition (VPR) was developed as an effective tool to improve mobile multimedia communication in ATS. VPR relates to the capability of a method or device to recognize and identify particular places or locations from the visual scene. This procedure involves inspecting visual data, like images or video frames, to control the unique features or features connected with diverse locations. By leveraging camera sensors, VPR allows vehicles to detect their surroundings, enabling context-aware communication and enhancing the entire system's performance. DL-empowered VPR offers a transformative manner to improve mobile multimedia communication in ATS. By identifying and understanding their situation, autonomous vehicles can communicate most effectively and operate reliably and safely, paving the way for a future characterized by seamless and intelligent transportation. This article develops a novel Deep Learning-Powered Visual Place Recognition for Enhanced Multimedia Communication in Autonomous Transport Systems (DLVPR-MCATS) methodology. The main aim of the DLVPR-MCATS methodology is to recognize visual places or not utilize optimal DL approaches. For this purpose, the DLVPR-MCATS approach utilizes a bilateral filtering (BF) based preprocessing model. For the feature fusion model, the DLVPR-MCATS approach follows three models: residual network (ResNet), EfficientNet, and MobileNetv2. Moreover, the hyperparameter tuning method uses the Harris Hawks Optimization (HHO) model. Finally, the bidirectional long short-term memory (BiLSTM) technique is implemented to recognize visual places. A wide range of simulations is executed to validate the solution of the DLVPR-MCATS method. The experimental validation of the DLVPR-MCATS method portrayed a superior performance over other models concerning various aspects.
visual place recognitionautonomous vehiclesmobile multimedia communicationautonomous transportation
Deep learning and artificial intelligence offer promising tools for improving the accuracy and efficiency of diagnosing various lung conditions using portable chest x-rays (CXRs). This study explores this potential by leveraging a large dataset containing over 6,000 CXR images from publicly available sources. These images encompass COVID-19 cases, normal cases, and patients with viral or bacterial pneumonia. The research proposes a novel approach called "Enhancing COVID Prediction with ESN-MDFS" that utilizes a combination of an Extreme Smart Network (ESN) and a Mean Dropout Feature Selection Technique (MDFS). This study aimed to enhance multi-class lung condition detection in portable chest X-rays by combining static texture features with dynamic deep learning features extracted from a pre-trained VGG-16 model. To optimize performance, preprocessing, data imbalance, and hyperparameter tuning were meticulously addressed. The proposed ESN-MDFS model achieved a peak accuracy of 96.18% with an AUC of 1.00 in a six-fold cross-validation. Our findings demonstrate the model's superior ability to differentiate between COVID-19, bacterial pneumonia, viral pneumonia, and normal conditions, promising significant advancements in diagnostic accuracy and efficiency.
Video surveillance systems have been instrumental in the consumer electronics industry, providing advanced functionalities such as product detection. This system utilizes sophisticated techniques to analyze real-time video feeds, enabling us to detect and track different objects, like products. By incorporating product detection abilities into video surveillance technology, consumers may improve security measures while gaining meaningful information in personalized advertising, inventory management, and retail analytics. This convergence of product detection and video surveillance opens new possibilities for businesses to optimize customer experiences, operations, and security. By leveraging deep learning (DL) for product detection, companies can provide enhanced consumer experiences, improve efficiency, and streamline operations. The incorporation of DL into video surveillance systems represented a breakthrough in the intersection of technology and security within the consumer electronics field. This study develops a new feature fusion technique with deep learning-based next-generation consumer products detection (FFDL-NGCPD) on video surveillance monitoring systems. The main aim of the FFDL-NGCPD system is to detect and classify consumer products. In the FFDL-NGCPD technique, a feature fusion process comprises two DL models: ShuffleNet and MobileNet. To improve the performance of the FFDL-NGCPD technique, the spider monkey optimization (SMO) algorithm is applied for the hyperparameter selection process. Finally, an extreme learning machine (ELM) classifier recognizes the various consumer products. The performance evaluation of the FFDL-NGCPD approach is verified utilizing a benchmark dataset. An extensive comprehensive study underlined the enhanced detection results of the FFDL-NGCPD technique.
video surveillancefeature fusionShuffleNetMobileNet
Detecting Alzheimer’s disease (AD) using magnetic resonance imaging (MRI) is essential for early diagnosis and management. This study introduces a new method for detecting AD by combining three robust models: DenseNet201, EfficientNet B7, and extremely randomized trees (ERT). We improve the ability to extract features in DenseNet201 by including a self-attention mechanism. Additionally, we use early stopping techniques on EfficientNet B7 to address the issue of overfitting. In addition, Bayesian Optimization and Hyperband optimization techniques are used to adjust the hyperparameters of extra-trees to differentiate normal and abnormal MRI images. In addition, the authors used SHapley Additive exPlanations to understand the model’s decision. With minimal computer resources, the proposed model achieved a remarkable accuracy of 98.9% in detecting AD. The findings highlight the effectiveness of recommended feature extraction and ERT models and optimization methods to accurately identify AD using MRI images.
Coronary artery disease (CAD) is one of the major causes of fatalities across the globe. The recent developments in convolutional neural networks (CNN) allow researchers to detect CAD from computed tomography (CT) images. The CAD detection model assists physicians in identifying cardiac disease at earlier stages. The recent CAD detection models demand a high computational cost and a more significant number of images. Therefore, this study intends to develop a CNN-based CAD detection model. The researchers apply an image enhancement technique to improve the CT image quality. The authors employed You look only once (YOLO) V7 for extracting the features. Aquila optimization is used for optimizing the hyperparameters of the UNet++ model to predict CAD. The proposed feature extraction technique and hyperparameter tuning approach reduces the computational costs and improves the performance of the UNet++ model. Two datasets are utilized for evaluating the performance of the proposed CAD detection model. The experimental outcomes suggest that the proposed method achieves an accuracy, recall, precision, F1-score, Matthews correlation coefficient, and Kappa of 99.4, 98.5, 98.65, 98.6, 95.35, and 95 and 99.5, 98.95, 98.95, 98.95, 96.35, and 96.25 for datasets 1 and 2, respectively. In addition, the proposed model outperforms the recent techniques by obtaining the area under the receiver operating characteristic and precision-recall curve of 0.97 and 0.95, and 0.96 and 0.94 for datasets 1 and 2, respectively. Moreover, the proposed model obtained a better confidence interval and standard deviation of [98.64-98.72] and 0.0014, and [97.41-97.49] and 0.0019 for datasets 1 and 2, respectively. The study's findings suggest that the proposed model can support physicians in identifying CAD with limited resources.
Although Kyphosis, an excessive forward rounding of the upper back, can occur at any age, adolescence is the most common time for Kyphosis. Surgery is frequently performed on Kyphosis patients; however, the condition may persist after the operation. The tricky part is figuring out, based on the patient’s traits, if the Kyphosis condition will continue after the treatment. There have been numerous models employed in the past to predict the Kyphosis disease, including Logistic Regression (LR), Naive Bayes (NB), Random Forest (RF), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Deep Neural Network (DNN), and others. Unfortunately, the precision was overestimated. Based on the dataset received from Kaggle, we investigated how to predict Kyphosis disorders more accurately by using these models with Hyperparameter tuning. While the calculations were being performed, certain variables were modified. The accuracy was increased by optimizing the fit parameters based on Hyperparameter tuning. Accuracy, recall or sensitivity, specificity, precision, balanced accuracy score, F1 score, and AUC-ROC score of all models, including the Hyperparameter tuning, were compared. Overall, the Hyperparameter-tuned DNN models excelled over the other models. The DNN models’ accuracy was 87.72% with 5-fold cross-validation and 87.64% with 10-fold cross-validation. It is advised that when a patient has a clinical procedure, the DNN model be trained to detect and foresee Kyphosis disease. Medical experts can use this study’s findings to correctly predict if a patient will still have Kyphosis after surgery. We propose that deep learning should be adopted and utilized as a crucial and necessary tool throughout the broad range of resolving biological queries.
Transactions on Emerging Telecommunications Technologies · vol. 34 (6)
Abstract Integrating physical, computational, and networking resources are the goal of cyber‐physical systems, also known as smart‐embedded systems. By investing in a solid foundation, we can improve the usefulness and timeliness of the services we rely on in every facet of our lives and ultimately live more elegantly. Regarding modern technology, data security is a significant factor that must be considered. The complexity of cyber‐physical systems' interacting components and middleware presents serious hurdles when it comes to protecting them from cyber‐attacks without negatively impacting their performance. This article proposes a unique, efficient encryption technique for anticipating cyber assaults in cyber‐physical systems, which addresses these concerns. The suggested method uses Bayesian optimization techniques to fine‐tune the LightGBM algorithm's hyper‐parameters. This proposed algorithm has been implemented on the intrusion detection dataset (UNR‐IDD) from the University of Nevada. Reno has been used to test the suggested approach. The proposed system achieved 99.17% accuracy, 0.9918 precision, and 0.9922 recall values. Our empirical evaluation demonstrates that the algorithm successfully increases accuracy and AUC value, making the cyber‐physical system more secure. In turn, the suggested methodology may offer robust assurance for user data safety.
Acute ischemic stroke (AIS) is a leading cause of mortality and disability. Over the past several decades, researchers proposed multiple techniques to enhance the effectiveness of AIS diagnosis. Magnetic resonance imaging (MRI) and computed tomography (CT) are widely used for assessing and treating AIS. In recent years, healthcare centers have applied the deep learning–based technique to support physicians in identifying diseases at earlier stages. Recently, researchers have employed convolutional neural network (CNN)-based image classifiers to detect diseases using complex images. However, the models demand high computation resources for generating a reasonable outcome. Thus, this study intends to build a fine-tuned CNN model for identifying AIS from MRI and CT images. The proposed framework contains three phases: image enhancement, feature extraction, and fine-tuned detection model. Initially, the researcher applies an image colorization technique using generative adversarial networks. You only look once V7 is used to extract the images’ features. In the second phase, the authors employed the Aquila optimization algorithm for tuning the hyperparameters of the Residual Network with Split attention (ResNest) model. To evaluate the performance of the ResNest model, the authors utilized two benchmark datasets of brain MRI and CT images. The MRI datasets contain 1021 healthy and 955 unhealthy images, whereas the CT datasets comprise 1551 healthy and 950 unhealthy images. The findings reveal that the ResNest model outperforms the recent approaches. The model achieves an average accuracy and F 1 score of 98.25 and 97.275, and 98.65 and 98.25, for the MRI and CT datasets, respectively. In addition, the ResNest model obtained a confidence interval score of [97.84-98.13] and [97.91-98.52] for the MRI and CT datasets, respectively. The study uniquely develops a lightweight application through a compelling data preprocessing and feature extraction technique. In addition, the fine-tuned ResNest model achieves a superior outcome with limited resources. Healthcare centers can implement this lightweight model for diagnosing AIS patients.
Red palm weevil (RPW) is a pest that can cause severe damage to plantations and affects palm trees. Classical approaches to detection depend on visual analysis, which is inaccurate and time-consuming. Hence, deep learning techniques emerge as a potential solution used for automating the process of detection and presenting efficient and precise results. The initial detection of the RPW remains a difficult task for good production as the identification will protect palm trees infected from the RPW. So advanced technologies like artificial intelligence (AI) and computer vision (CV) can be used in preventing the spread of the RPW on palm trees. Various scholars still working on identifying a precise method for the classification, identification, and localization of the RPW pest. This article develops an automated Red Palm Weevil Detection using Gorilla Troops Optimizer with Deep Learning (RPWD-GTODL) method. The goal of the presented RPWD-GTODL approach lies in the accurate detection and localization of the RPW effectually. To accomplish this, the presented RPWD-GTODL technique initially uses the Gabor filtering (GF) technique to pre-process the images. For RPW detection, the RPWD-GTODL technique uses a Mask RCNN object detector with MobileNetv2 as a backbone network. Moreover, the detection performance of the RPWD-GTODL technique can be boosted by the design of the GTO algorithm for the hyperparameter selection of the MobileNetv2 model. The performance validation of the RPWD-GTODL technique is tested using the RPW dataset and the results demonstrate the enhanced performance on RPW detection process with maximum accuracy of 99.27%.
International Journal of Advanced Computer Science and Applications · vol. 14 (2)
A convolutional neural network (CNN) is a subset of machine learning as well as one of the different types of artificial neural networks that are used for different applications and data types. Activation functions (AFs) are used in this type of network to determine whether or not its neurons are activated. One non-linear AF named as Rectified Linear Units (ReLU) which involves a simple mathematical operations and it gives better performance. It avoids rectifying vanishing gradient problem that inherents older AFs like tanh and sigmoid. Additionally, it has less computational cost. Despite these advantages, it suffers from a problem called Dying problem. Several modifications have been appeared to address this problem, for example; Leaky ReLU (LReLU). The main concept of our algorithm is to improve the current LReLU activation functions in mitigating the dying problem on deep learning by using the readjustment of values (changing and decreasing value) of the loss function or cost function while number of epochs are increased. The model was trained on the MNIST dataset with 20 epochs and achieved lowest misclassification rate by 1.2%. While optimizing our proposed methods, we received comparatively better results in terms of simplicity, low computational cost, and with no hyperparameters.