Advancing Phishing Email Detection: A Comparative Study of Deep Learning Models.

Phishing is one of the most dangerous attacks targeting individuals, organizations, and nations. Although many traditional methods for email phishing detection exist, there is a need to improve accuracy and reduce false-positive rates. Our work investigates one-dimensional CNN-based models (1D-CNNPD) to detect phishing emails in order to address these challenges. Additionally, further improvement is achieved with the augmentation of the base 1D-CNNPD model with recurrent layers, namely, LSTM, Bi-LSTM, GRU, and Bi-GRU, and experimented with the four resulting models. Two benchmark datasets were used to evaluate the performance of our models: Phishing Corpus and Spam Assassin. Our results indicate that, in general, the augmentations improve the performance of the 1D-CNNPD base model. Specifically, the 1D-CNNPD with Bi-GRU yields the best results. Overall, the performance of our models is comparable to the state of the art of CNN-based phishing email detection. The Advanced 1D-CNNPD with Leaky ReLU and Bi-GRU achieved 100% precision, 99.68% accuracy, an F1 score of 99.66%, and a recall of 99.32%. We observe that increasing model depth typically leads to an initial performance improvement, succeeded by a decline. In conclusion, this study highlights the effectiveness of augmented 1D-CNNPD models in detecting phishing emails with improved accuracy. The reported performance measure values indicate the potential of these models in advancing the implementation of cybersecurity solutions to combat email phishing attacks.

Uncovering hidden genetic risk factors for breast and ovarian cancers in BRCA-negative women: a machine learning approach in the Saudi population.

Breast and ovarian cancers are prevalent worldwide, with genetic factors such as BRCA1 and BRCA2 mutations playing a significant role. However, not all patients carry these mutations, making it challenging to identify risk factors. Researchers have turned to whole exome sequencing (WES) as a tool to identify genetic risk factors in BRCA-negative women. WES allows the sequencing of all protein-coding regions of an individual's genome, providing a comprehensive analysis that surpasses traditional gene-by-gene sequencing methods. This technology offers efficiency, cost-effectiveness and the potential to identify new genetic variants contributing to the susceptibility to the diseases. Interpreting WES data for disease-causing variants is challenging due to its complex nature. Machine learning techniques can uncover hidden genetic-variant patterns associated with cancer susceptibility. In this study, we used the extreme gradient boosting (XGBoost) and random forest (RF) algorithms to identify BRCA-related cancer high-risk genes specifically in the Saudi population. The experimental results exposed that the RF method scored superior performance with an accuracy of 88.16% and an area under the receiver-operator characteristic curve of 0.95. Using bioinformatics analysis tools, we explored the top features of the high-accuracy machine learning model that we built to enhance our knowledge of genetic interactions and find complex genetic patterns connected to the development of BRCA-related cancers. We were able to identify the significance of HLA gene variations in these WES datasets for BRCA-related patients. We find that immune response mechanisms play a major role in the development of BRCA-related cancer. It specifically highlights genes associated with antigen processing and presentation, such as HLA-B, HLA-A and HLA-DRB1 and their possible effects on tumour progression and immune evasion. In summary, by utilizing machine learning approaches, we have the potential to aid in the development of precision medicine approaches for early detection and personalized treatment strategies.

Prevalence and Pattern of Injuries Across the Weight-Training Sports.

Background The prevalence and pattern of injury among weightlifters are insufficiently documented despite these research works. Understanding the injury pattern is crucial for minimizing side effects and maximizing the advantages of weight training. Therefore, the purpose of this study is to determine the frequency and pattern of musculoskeletal injury among weightlifters and to investigate the associations between the prevalence of injury and sociodemographic and training characteristics variables. Methods A descriptive cross-sectional, questionnaire-based study was conducted. An online questionnaire was designed by Google Forms to collect the data by using a self-administered questionnaire. From all health clubs in Taif city, Saudi Arabia, one club was chosen by simple random sampling methodology, where all attendant weightlifters during the study period were contacted to participate in the study. Data was entered on the computer using Microsoft Office Excel 2016 for Windows. Qualitative data was expressed as numbers and percentages, and the Chi-squared test (χ2) was used to assess the relationship between variables. A p-value < 0.05 was considered statistically significant. Results The study included 393 participants, and most respondents fall within the age range of 18-29, accounting for 60.1% of the total. About 27% of participants had a weightlift injury during the last six months. The body parts most injured during weightlifting include the shoulder (7.4%), knee (4.6%), and wrist (3.6%). In terms of the type of injuries sustained, inflammation and pain in the bending of the body (5.9%), torsion (3.6%), ligament tear/muscle tear (3.8%), and stripped-off injuries (2.3%) were reported. Conclusion Musculoskeletal injuries are prevalent among weightlifters due to the nature of the sport and the demands it places on the body. There was no significant association between the injury occurrence with gender, age, or body mass index. However, there was a significant association between the occurrence of injury and weight carried while lifting weights.

A semi supervised approach to Arabic aspect category detection using Bert and teacher-student model.

Aspect-based sentiment analysis tasks are well researched in English. However, we find such research lacking in the context of the Arabic language, especially with reference to aspect category detection. Most of this research is focusing on supervised machine learning methods that require the use of large, labeled datasets. Therefore, the aim of this research is to implement a semi-supervised self-training approach which utilizes a noisy student framework to enhance the capability of a deep learning model, AraBERT v02. The objective is to perform aspect category detection on both the SemEval 2016 hotel review dataset and the Hotel Arabic-Reviews Dataset (HARD) 2016. The four-step framework firstly entails developing a teacher model that is trained on the aspect categories of the SemEval 2016 labeled dataset. Secondly, it generates pseudo labels for the unlabeled HARD dataset based on the teacher model. Thirdly, it creates a noisy student model that is trained on the combined datasets (∼1 million sentences). The aim is to minimize the combined cross entropy loss. Fourthly, an ensembling of both teacher and student models is carried out to enhance the performance of AraBERT. Findings indicate that the ensembled teacher-student model demonstrates a 0.3% improvement in its micro F1 over the initial noisy student implementation, both in predicting the Aspect Categories in the combined datasets. However, it has achieved a 1% increase over the micro F1 of the teacher model. These results outperform both baselines and other deep learning models discussed in the related literature.

Cost effectiveness of rivaroxaban versus warfarin among nonvalvular atrial fibrillation patients in Saudi Arabia: A Single-Center retrospective cohort study.

Background: Rivaroxaban is a novel oral anticoagulant (NOAC) that is commonly used for stroke prevention among patients with atrial fibrillation (AF). However, its cost effectiveness in reducing the risk of hospitalization and mortality in comparison to warfarin among nonvalvular AF patients in Saudi Arabia is largely unknown. Methods: This was a single-center retrospective chart review of adult patients (≥18 years) with nonvalvular AF who were treated with warfarin or rivaroxaban for at least 12 months. Patients with mitral valve stenosis were excluded from the study. Multiple logistic regression was conducted to examine the risk of hospitalization and mortality as a composite outcome, and all annual healthcare costs were captured. Inverse probability treatment weighting with bootstrapping was conducted to determine the mean costs and effectiveness rates. Results: Two-hundred and twenty-six patients (142 on rivaroxaban and 84 on warfarin) met the inclusion criteria and were included in the analysis. Most of the patients were females (65.91 %), had diabetes (50.57 %) and hypertension (73.76 %), and with a mean age of 68.95 ± 12.55 years. No significant difference in the odds of the composite outcome for rivaroxaban versus warfarin was found (OR = 0.785, 95 % CI = [0.427-1.446], p = 0.443). Rivaroxaban resulted in a mean annual cost saving of $13,260.79 with an 87.65 % confidence level that it would be more effective than warfarin with a mean difference in effectiveness rate of 0.168 % (95 % CI [-5.210-18.36]). Conclusion: Rivaroxaban was associated with lower direct medical costs and non-inferior effectiveness among nonvalvular AF patients in comparison to warfarin.

Normative values and factors affecting Pediatric Reach Tests in Saudi children aged 6-11 years in the eastern province: cross-sectional study.

Background: The Pediatric Reach Tests (PRTs) assess balance while standing-the Functional Reach Test (FRT) and Lateral Reach Test (LRT)-and in a sitting position-the Modified Functional Reach Test (MFRT) and Modified Lateral Reach Test (MLRT). Normative values have not been fully evaluated in Saudi children. The objectives are; to estimate the normative values for PRTs; investigate the correlation between the PRTs and demographic/anthropometric characteristics; and develop predictive equations for the PRTs. Methods: In this cross-sectional study, 251 children aged 6-11 were recruited. The PRTs were measured and correlated with demographic/anthropometric variables. A stepwise regression was conducted to develop the predictive equations for the PRT scores. Results: The mean and standard deviations (in cm) of the PRT scores were as follows: FRT = 20.02 ± 4.31; LRT = 13.42 ± 3.38; MFRT = 21.49 ± 4.70, and MLRT = 14.64 ± 3.66. Several significant correlations were found. Moderate correlations existed between the PRT scores and age, height, upper extremity length, lower extremity length, and foot length; there was a weak correlation with body mass index. Weight was moderately correlated with FRT and MFRT and weakly correlated with LRT and MLRT. The correlation between the base of support and LRT was moderate and was weak with FRT, MFRT, and MLRT. A weak correlation was found between sex and LRT. Age and height were the most predictive of PRT scores. Conclusion: This study provided PRT normative values that can be used as a clinical reference for evaluating balance in typically developing children.

Effect of surgical face mask wearing on tear film in women with a high body mass index.

PURPOSE: Due to the COVID-19 pandemic, wearing a face mask has become an essential measure to reduce the rate of virus spreading. The aim of the study was to assess the effect of wearing a surgical face mask for a short period on the tear film parameters in subjects with a high body mass index (BMI). METHODS: Twenty-five females with a high BMI (31.4 ± 5.5 kg/m2) aged 18-35 years (22.7 ± 4.6 years) participated in the study. In addition, a control group consisting of 25 females (23.0 ± 6.7 years) with a high BMI (29.9 ± 4.1 kg/m2) participated in the study in which no mask was worn. The standardized patient evaluation of eye dryness (SPEED) questionnaire was completed first, followed by the phenol red thread (PRT) and tear ferning (TF) tests, before wearing the face mask. The subjects wore the face mask for 1 hour, and the measurements were performed again immediately after its removal. For the control group, the measurements were performed twice with one hour gap. RESULTS: Significant (Wilcoxon test, p < 0.05) differences were found between the SPEED scores (p = 0.035) and the PRT measurement (p = 0.042), before and after wearing the surgical face mask. The PRT scores have improved after wearing the surgical face mask, while the dry eye symptoms detected by the SPEED questionnaire have increased. On the other hand, no significant (Wilcoxon test, p = 0.201) differences were found between the TF grades before and after wearing a surgical face mask. For the control group, no significant (Wilcoxon test, p > 0.05) differences were found between the two scores from the SPEED questionnaire and the PRT, and TF tests. CONCLUSIONS: Wearing a surgical face mask for a short duration leads to a change in volume and quality of tears as well as dry eye symptoms in women with a high BMI.

Adversarial attacks against supervised machine learning based network intrusion detection systems.

Adversarial machine learning is a recent area of study that explores both adversarial attack strategy and detection systems of adversarial attacks, which are inputs specially crafted to outwit the classification of detection systems or disrupt the training process of detection systems. In this research, we performed two adversarial attack scenarios, we used a Generative Adversarial Network (GAN) to generate synthetic intrusion traffic to test the influence of these attacks on the accuracy of machine learning-based Intrusion Detection Systems(IDSs). We conducted two experiments on adversarial attacks including poisoning and evasion attacks on two different types of machine learning models: Decision Tree and Logistic Regression. The performance of implemented adversarial attack scenarios was evaluated using the CICIDS2017 dataset. Also, it was based on a comparison of the accuracy of machine learning-based IDS before and after attacks. The results show that the proposed evasion attacks reduced the testing accuracy of both network intrusion detection systems models (NIDS). That illustrates our evasion attack scenario negatively affected the accuracy of machine learning-based network intrusion detection systems, whereas the decision tree model was more affected than logistic regression. Furthermore, our poisoning attack scenario disrupted the training process of machine learning-based NIDS, whereas the logistic regression model was more affected than the decision tree.

Location-Allocation Model to Improve the Distribution of COVID-19 Vaccine Centers in Jeddah City, Saudi Arabia.

The correct distribution of service facilities can help keep fixed and overhead costs low while increasing accessibility. When an appropriate location is chosen, public-sector facilities, such as COVID-19 centers, can save lives faster and provide high-quality service to the community at a low cost. The purpose of the research is to highlight the issues related to the location of COVID-19 vaccine centers in the city of Jeddah, Saudi Arabia. In particular, this paper aims to analyze the accessibility of COVID-19 vaccine centers in Jeddah city using maximal coverage location problems with and without constraint on the number and capacity of facilities. A maximal coverage model is first used to analyze the COVID-19 vaccination coverage of Jeddah districts with no restriction on the facility capacity. Then, a maximize capacitated coverage method is utilized to assess the centers' distribution and demand coverage with capacity constraints. Finally, the minimize facilities model is used to identify the most optimal location required to satisfy all demand points with the least number of facilities. The optimization approaches consider the objective function of minimizing the overall transportation time and travel distance to reduce wastage on the service rate provided to the patients. The optimization model is applied to a real-world case study in the context of the COVID-19 vaccination center in Jeddah. The results of this study provide valuable information that can help decision-makers locate and relocate COVID-19 centers more effectively under different constraints conditions.

A Survey of Location-Allocation of Points of Dispensing During Public Health Emergencies.

Public health emergencies such as disease outbreaks and bioterrorism attacks require immediate response to ensure the safety and well-being of the affected community and prevent the further spread of infection. The standard method to increase the efficiency of mass dispensing during health emergencies is to create emergency points called points of dispensing (PODs). PODs are sites for distributing medical services such as vaccines or drugs to the affected population within a specific time constraint. These PODs need to be sited in optimal locations and have people (demand points) assigned to them simultaneously; this is known as the location-allocation problem. PODs may need to be selected to serve the entire population (full allocation) or different priority or needs groups (partial allocation). Several previous studies have focused on location problems in different application domains, including healthcare. However, some of these studies focused on healthcare facility location problems without specifying location-allocation problems or the exact domain. This study presents a survey of the PODs location-allocation problem during public health emergencies. This survey aims to review and analyse the existing models for PODs location-allocation during public health emergencies based on full and partial demand points allocation. Moreover, it compares existing models based on their key features, strengths, and limitations. The challenges and future research directions for PODs location-allocation models are also discussed. The results of this survey demonstrated a necessity to develop a variety of techniques to analyse, define and meet the demand of particular groups. It also proved essential that models be developed for different countries, including accounting for variations in population size and density. Moreover, the model constraints, such as those relating to time or prioritizing certain groups, need to be considered in the solution. Finally, additional comparative studies are required to clarify which methods or models are adequate based on predefined criteria.

Spatial Analysis of COVID-19 Vaccine Centers Distribution: A Case Study of the City of Jeddah, Saudi Arabia.

The COVID-19 pandemic is one of the most devastating public health emergencies in history. In late 2020 and after almost a year from the initial outbreak of the novel coronavirus (SARS-CoV-2), several vaccines were approved and administered in most countries. Saudi Arabia has established COVID-19 vaccination centers in all regions. Various facilities were selected to set up these vaccination centers, including conference and exhibition centers, old airport terminals, pre-existing medical facilities, and primary healthcare centers. Deciding the number and locations of these facilities is a fundamental objective for successful epidemic responses to ensure the delivery of vaccines and other health services to the entire population. This study analyzed the spatial distribution of COVID-19 vaccination centers in Jeddah, a major city in Saudi Arabia, by using GIS tools and methods to provide insight on the effectiveness of the selection and distribution of the COVID-19 vaccination centers in terms of accessibility and coverage. Based on a spatial analysis of vaccine centers' coverage in 2020 and 2021 in Jeddah presented in this study, coverage deficiency would have been addressed earlier if the applied GIS analysis methods had been used by authorities while gradually increasing the number of vaccination centers. This study recommends that the Ministry of Health in Saudi Arabia evaluated the assigned vaccination centers to include the less-populated regions and to ensure equity and fairness in vaccine distribution. Adding more vaccine centers or reallocating some existing centers in the denser districts to increase the coverage in the uncovered sparse regions in Jeddah is also recommended. The methods applied in this study could be part of a strategic vaccination administration program for future public health emergencies and other vaccination campaigns.

Time series predicting of COVID-19 based on deep learning.

COVID-19 was declared a global pandemic by the World Health Organisation (WHO) on 11th March 2020. Many researchers have, in the past, attempted to predict a COVID outbreak and its effect. Some have regarded time-series variables as primary factors which can affect the onset of infectious diseases like influenza and severe acute respiratory syndrome (SARS). In this study, we have used public datasets provided by the European Centre for Disease Prevention and Control for developing a prediction model for the spread of the COVID-19 outbreak to and throughout Malaysia, Morocco and Saudi Arabia. We have made use of certain effective deep learning (DL) models for this purpose. We assessed some specific major features for predicting the trend of the existing COVID-19 outbreak in these three countries. In this study, we also proposed a DL approach that includes recurrent neural network (RNN) and long short-term memory (LSTM) networks for predicting the probable numbers of COVID-19 cases. The LSTM models showed a 98.58% precision accuracy while the RNN models showed a 93.45% precision accuracy. Also, this study compared the number of coronavirus cases and the number of resulting deaths in Malaysia, Morocco and Saudi Arabia. Thereafter, we predicted the number of confirmed COVID-19 cases and deaths for a subsequent seven days. In this study, we presented their predictions using the data that was available up to December 3rd, 2020.

On the prediction of isolation, release, and decease states for COVID-19 patients: A case study in South Korea.

A respiratory syndrome COVID-19 pandemic has become a serious public health issue nowadays. The COVID-19 virus has been affecting tens of millions people worldwide. Some of them have recovered and have been released. Others have been isolated and few others have been unfortunately deceased. In this paper, we apply and compare different machine learning approaches such as decision tree models, random forest, and multinomial logistic regression to predict isolation, release, and decease states for COVID-19 patients in South Korea. The prediction can help health providers and decision makers to distinguish the states of infected patients based on their features in early intervention to take an action either by releasing or isolating the patient after the infection. The proposed approaches are evaluated using Data Science for COVID-19 (DS4C) dataset. An analysis of DS4C dataset is also provided. Experimental results and evaluation show that multinomial logistic regression outperforms other approaches with 95% in a state prediction accuracy and a weighted average F1-score of 95%.

An End-to-End Human Abnormal Behavior Recognition Framework for Crowds With Mentally Disordered Individuals.

Abnormal or violent behavior by people with mental disorders is common. When individuals with mental disorders exhibit abnormal behavior in public places, they may cause physical and mental harm to others as well as to themselves. Thus, it is necessary to monitor their behavior using visual surveillance systems. However, it is challenging to automatically detect human abnormal behavior (especially for individuals with mental disorders) based on motion recognition technologies. To address these issues, in the current work, we propose an end-to-end abnormal behaviour detection framework from a new perspective in conjunction with the Graph Convolutional Network (GCN) and a 3D Convolutional Neural Network (3DCNN). Specifically, we first train a one-class classifier to extract features and estimate abnormality scores. To improve the performance of abnormal behavior detection, GCN is used to model the similarity between video clips for the correction of noisy labels. Then, based on this framework, GCN recognizes the normal behavior clips in the abnormal video and removes them, while the clips identified as abnormal behavior are retained. Finally, a 3D CNN is used to extract spatiotemporal features to classify different abnormal behaviors. In order to better detect the violent behavior of individuals with mental disorders, the paper focuses on the UCF-Crime dataset with various types of violent behaviors. By experimenting with this dataset, the classification accuracy reaches 37.9%, which is significantly better than that of the current state-of-the-art approaches.

Attractivity criterion on a delayed tick population dynamics equation with a reproductive function $ f(u) = ru

The aim of this article is to analyze the delay influence on the attraction for a scalar tick population dynamics equation accompanying two disparate delays. Taking advantage of the fluctuation lemma and some dynamic inequalities, we derive a criterion to assure the persistence and positiveness on the considered model. Furthermore, a time-lag-dependent condition is proposed to insure the global attractivity for the addressed model. Besides, we give some simulation diagrams to substantiate the validity of the theoretical outcomes.

Cardiovascular Disease Recognition Based on Heartbeat Segmentation and Selection Process.

Assessment of heart sounds which are generated by the beating heart and the resultant blood flow through it provides a valuable tool for cardiovascular disease (CVD) diagnostics. The cardiac auscultation using the classical stethoscope phonological cardiogram is known as the most famous exam method to detect heart anomalies. This exam requires a qualified cardiologist, who relies on the cardiac cycle vibration sound (heart muscle contractions and valves closure) to detect abnormalities in the heart during the pumping action. Phonocardiogram (PCG) signal represents the recording of sounds and murmurs resulting from the heart auscultation, typically with a stethoscope, as a part of medical diagnosis. For the sake of helping physicians in a clinical environment, a range of artificial intelligence methods was proposed to automatically analyze PCG signal to help in the preliminary diagnosis of different heart diseases. The aim of this research paper is providing an accurate CVD recognition model based on unsupervised and supervised machine learning methods relayed on convolutional neural network (CNN). The proposed approach is evaluated on heart sound signals from the well-known, publicly available PASCAL and PhysioNet datasets. Experimental results show that the heart cycle segmentation and segment selection processes have a direct impact on the validation accuracy, sensitivity (TPR), precision (PPV), and specificity (TNR). Based on PASCAL dataset, we obtained encouraging classification results with overall accuracy 0.87, overall precision 0.81, and overall sensitivity 0.83. Concerning Micro classification results, we obtained Micro accuracy 0.91, Micro sensitivity 0.83, Micro precision 0.84, and Micro specificity 0.92. Using PhysioNet dataset, we achieved very good results: 0.97 accuracy, 0.946 sensitivity, 0.944 precision, and 0.946 specificity.

An analytical study on the awareness, attitude and practice during the COVID-19 pandemic in Riyadh, Saudi Arabia.

BACKGROUND: Coronavirus disease (COVID-19) is an infectious disease caused by a new variable of the Coronaviridae family. COVID-19 spreads primarily by contacting the virus either from a COVID-19-infected individual through coughing or sneezing or from COVID-19-contaminated surfaces. On March 12, 2020, the World Health Organization (WHO) announced COVID-19 as a pandemic. The government of Saudi Arabia was among the first countries in the world to take quick and serious precautions. The Ministry of Health (MOH) has made the public aware of the virus transmission patterns and the importance of quarantine and curfew. Despite strict measures taken, the awareness of people towards infectious viruses remains the most important factor in limiting the widespread of diseases. METHOD: A cross-sectional survey of 1767 participants, was conducted to explore the awareness, attitude and practice of COVID-19 in relation to socioeconomic data among residents in the city of Riyadh. RESULTS: Of all the participants, 58% showed a moderate level of awareness, 95% presented a high attitude and 81% presented an adequate practice regarding COVID-19. Significant positive correlation between awareness-attitude (r = 0.132, p-value < 0.001) and attitude-practice (r = 0.149, p-value < 0.001) were found. The gender of the participants was the only common characteristic significantly associated with both awareness and practice. This study revealed that males showed a slight increase (60%) in the level of awareness compared to female participants (57%), however, when it comes to the practice towards COVID-19, females showed slightly better practice (82%) than males (80%). The World health organization (WHO) and the Ministry of Health (MOH) were the main sources of information. CONCLUSION: Despite the moderate public awareness, their attitude and practice were better. Therefore, public awareness must be improved to be prepared for epidemic and pandemic situations. A comprehensive public health education program is important to increase awareness and to reach sufficient knowledge.

Controlling COVID-19 Pandemic: A Mass Screening Experience in Saudi Arabia.

A highly accelerating number of people around the world have been infected with novel Coronavirus disease 2019 (COVID-19). Mass screening programs were suggested by the World Health Organization (WHO) as an effective precautionary measure to contain the spread of the virus. On 16 April 2020, a COVID-19 mass screening program was initiated in Saudi Arabia in multiple phases. This study aims to analyze the number of detected COVID-19 cases, their demographic data, and regions most affected in the initial two phases of these mass screening programs. A retrospective cross-sectional study was conducted among the high-risk population as part of the COVID-19 mass screening program across all regions in Saudi Arabia during April and May 2020. A Chi-square-test was used to determine the associations between positive cases and various demographic variables. Out of 71,854 screened individuals, 13.50% (n = 9701) were COVID-19 positive, of which 83.27% (n = 59,835) were males. Among positive cases, in the 30-39 years age group, 6.36% were in the active phase, and 2.19% were in the community phase. Based on our experience, launching mass screening programs is crucial for early case detection, isolation, and pattern recognition for immediate public interventions.