Accuration of Classification of Covid with Convolutional Neural Network-Based Image Chest X-ray with Variations in Image Size and Batch Size

The use of convolutional neural networks in Covid classification has a positive impact on the speed of justification and can provide high accuracy. But on the one hand, the many parameters on CNN will also have an impact on the resulting accuracy. CNN requires time and a heavy level of computation. Setting the right parameters will provide high accuracy. This study examines the performance of CNN with variations in image size and minibatch. Parameter settings used are max epoch values of 100, minibatch variations of 32, 64, and 128, and learning rate of 0.1 with image size inputs of 50,100, and 150 variations on the level of accuracy. The dataset consists of training data and test data, 200 images, which are divided into two categories of normal and abnormal images (Covid). The results showed an accuracy with the use of minibatch 128 with the highest level of accuracy at image size 150 × 150 on test data of 99,08%. The size of the input matrix does not always have an impact on increasing the level of accuracy, especially on the minibatch 32. The parameter setting on CNN was dependent on the CNN architecture, the dataset used, and the size of the dataset. One can imply that optimization parameter in CNN can approve good accuration. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Outdoor Social Distancing Violation System Detection Using YOLO Algorithm

The government issued orders to implement social distancing or physical distancing. Social distancing is a method of maintaining a distance of at least one meter from other people. This is useful for reducing/preventing disease transmission (virus) and reducing the chain of the spread of covid-19. So the hospital can provide optimal service. For this reason, this research is structured to create a system that can detect violations of social distancing in an open place. This system uses the You Only Look Once (YOLO) algorithm. The developed system uses a pre-Trained Yolov4 model to detect 80 object classes. Testing of this system is carried out based on several scenarios. The system is programmed using Python, with tools for coding Microsoft Visual Studio Code and Anaconda. The best result from creating the detection mode is obtained from a dataset ratio of 90% train data and 10% test data, with the mean average precision results obtained being 54.11%. © 2022 IEEE.

Analysis Pre and Post COVID-19 Pandemic Rorschach Test Data of Using EM Algorithms and GMM Models

The global spread of the COVID-19 virus has become one of the greatest challenges that humanity has faced in recent years. The unprecedented circumstances of forced isolation and uncertainty that it has imposed on us continue to impact our mental well-being, whether or not we have been directly affected by the virus. Over a period of nearly three years (2017-2020), data was collected from multiple administrations of the Rorschach test, one of the most renowned and extensively studied psychological tests. This study involved the clustering of data, collected through the RAP3 software, to analyze the distinctive trends in data recorded before and after the pandemic. This was achieved through the implementation of the well-established machine learning algorithm, Expectation-Maximization. The proposed solution effectively identifies the key variables that significantly influence the subject's score and provides a reliable solution. Additionally, the solution offers an intuitive visualization that can assist psychologists in accurately interpreting shifts in trends and response distributions within a large amount of data in the two periods. © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)

Computer-Aided Diagnosis of COVID-19 with Joint Instance Segmentation and Classification

To address the difficulty of small lesion area detection of COVID-19 patients in their lung CT images, the author has proposed an end-to-end network which using semantic segmentation to guide instance segmentation, and extending transfer learning to the classification of COVID-19 pneumonia, Common pneumonia and Normal. Firstly, in order to extract richer multi-scale features and increase the weight of low-level features, we have introduced the Atrous Spatial Pyramid Pooling(ASPP) into the Feature Pyramid Network(FPN), and proposed Multi-scale Reverse Attention Feature Pyramid Network, then having added a semantic segmentation branch to guide instance segmentation after the output of ASPP, finally, we have extracted the object category score by detector for auxiliary classification. Segmentation experiments were carried out on the dataset of CC-CCII and COVID-19 infection segmentation dataset, the mean average precision(mAP) is 39.57%, 35.36%, Compared with the COVID-CT-Mask-Net, it has improved by 5.52%, 2.33%, we also carried out classification experiments on the dataset that is from COVIDX-CT, the sensitivity and specificity of the model for detecting COVID-19 in test data are 95.88% and 98.95% respectively. Also, the sensitivity and specificity of the model for detecting Common pneumonia in test data are 98.62% and 99.25% respectively, the sensitivity and specificity of the model for detecting Normal in test data are 99.61% and 99.11% respectively, which are the best results based on this dataset and indicators, this shows that the proposed method can quickly and effectively help the clinician identify and diagnose COVID-19 patient through their lung CT images. © 2022 IEEE.

On forecasting of COVID-19 transmission in Saudi Arabia and Egypt using reservoir computing model

This work is devoted to introduce a reliable, fast and highly accurate reservoir computer machine learning scheme to forecast time evolution of COVID-19 pandemic. In particular, the COVID-19 official data related to susceptible cases, confirmed cases, and recovered cases in Egypt and Saudi Arabia are collected. They employed as the training data for suggested reservoir computer (RC) model. Then, detailed simulation experiments are carried out within specified time periods. The evolution of COVID-19 in Egypt and Saudi Arabia are predicted on the subsequent times intervals and compared with real validation test data. The forecasting accuracy is improved by computing the optimal output matrix which minimizes the normalized root mean square errors (NRMSEs). The performance of RC scheme is evaluated when different-size training data, different-size test data, and different number of internal nodes are used. The comparisons with the robust LSTM deep learning techniques are performed. It is shown that the presented RC-based forecasting technique is more accurate for long-time forecasting, faster, and has lower computational cost. © 2023 THE AUTHORS

Forecasting Covid-19 Time Series Data using the Long Short-Term Memory (LSTM)

Confirmed statistical data of Covid-19 cases that have accumulated sourced from (https://corona.riau.go.id/data-statistik/) in Riau Province on June 7, 2021, there were 63441 cases, on June 14, 2021, it increased to 65883 cases, on June 21, 2021, it increased to 67910, and on June 28, 2021, it increased to 69830 cases. Since the beginning of this pandemic outbreak, it has been observed that the case data continues to increase every week until this July. This study predicts cases of Covid-19 time series data in Riau Province using the LSTM algorithm, with a dataset of 64 lines. Long-Short Term Memory has the ability to store memory information for patterns in the data for a long time at the same time. Tests predicting historical data for Covid-19 cases in Riau Province resulted in the lowest RMSE value in the training data, which was 8.87, and the test data, which was 13.00, in the death column. The evaluation of the best MAPE value in the training data, which is 0.23%, is in the recovered column, and the evaluation of the best MAPE value in the test data, which is 0.27%, in the positive_number column. In the test to predict the next 30 days using the LSTM model that has been trained, it was found that the performance evaluation of the prediction results for the positive_number column and the death column was very good, the recovery column was categorized as good, the independent_isolation column and the care_rs column were categorized as poor. © 2022, International Journal of Advanced Computer Science and Applications. All Rights Reserved.

ULD-NUIG at Social Media Mining for Health Applications (#SMM4H) Shared Task 2021

In this paper, we present the ULD-NUIG team's system, designed as part of Social Media Mining for Health Applications (#SMM4H) Shared Task 2021. We participate in two tasks out of eight, namely "Classification of tweets self-reporting potential cases of COVID-19" (Task 5) and "Classification of COVID19 tweets containing symptoms" (Task 6). The team conduct a series of experiments to explore the challenges of both the tasks. We used a multilingual pre-trained BERT model for Task 5 and Generative Morphemes with Attention (GenMA) model for Task 6. In the experiments, we find that, GenMA, developed for Task 6, gives better results on both validation and test data-set. The submitted systems achieve F-1 score 0.53 for Task 5 and 0.84 for Task 6 on test data-set. © 2021 Association for Computational Linguistics.

Don’t Just Drop Them: Function Words as Features in COVID-19 Related Fake News Classification on Twitter

This research shows that function words can be useful as features for machine learning models tasked with detecting conspiratorial content in COVID-19 related Twitter posts. A significance test exposes that the distribution of function words between fake and legitimate content varies greatly. Further, a support vector machine classifier is demonstrated to perform above chance when using function word-only features, achieving a Matthews correlation coefficient of 0.139 on unseen test data. Copyright 2021 for this paper by its authors.

The Effectiveness of ScratchJr in Enhancing Preschool Children's Problem-Solving Skills

Utilizing digital media serves as one of the teachers' efforts to develop children's problem-solving skills. By using the ScratchJr program, teachers can develop different play activities to stimulate children's problem-solving skills. This study sought to examine the effectiveness of the ScratchJr computer program in enhancing 5-6 years old children's skills during the Covid-19 pandemic. To this end, a quasi-experimental design was applied. The study was conducted in an Islamic kindergarten in Surakarta by assigning 29 students to an experimental group and another 29 students to a control group. The data were collected using test and nontest techniques. The test-based data collection was done using pre-and post-tests, while the non-test data collection was carried out through observation. The hypothesis was tested using parametric statistics because the data were considered to be normally distributed and homogeneous. An independent sample t-test was applied to analyze the data. The result indicates that the ScratchJr program effectively enhances 5-6 years old children's problem-solving skills, as shown by the significant difference between the score of experimental and control groups (ρ < 0.05). © 2021 ACM.

Binary Logistic Regression and Linear Discriminant Analyses in Evaluating Laboratory Factors Associated with COVID-19: A Comparison of Two Statistical Methods

Introduction: Few studies have been conducted to construct a reliable predictive model for the differential diagnosis of severe and non-severe Coronavirus disease-2019 (COVID-19) in the early stages of the disease. This study aimed to compare the accuracy of linear discriminate analysis (LDA) and binary logistic regression (BLR), as two empirical correlations, in predicting COVID-19 severity using single laboratory data and calculated indexes such as the neutrophil-to-lymphocyte ratio (NLR) and systemic immune-inflammation index (SII). Materials and Methods: We investigated 109 patients with confirmed COVID-19 pneumonia. Epidemiological, demographic, clinical, laboratory, and outcome data were obtained, and the patients were classified into two groups: mild group (42 patients) and severe group (67 patients). Results: A comparison of the clinical data in the severe and non-severe groups showed significant differences in SpO(2) and respiratory rate. In addition, significant difference in NLR, SII, white blood cell count, neutrophil count, mean corpuscular volume and mean corpuscular hemoglobin, lymphocyte count, erythrocyte sedimentation rate, lactate dehydrogenase, and blood urea nitrogen was found between both groups. Moreover, there was a small difference between the LDA and LR models, and LDA was more appropriate for a smaller sample size. Conclusion: Our predictive models could help clinicians to identify patients at risk of severe COVID-19 Such prediction can be performed by a simple blood test. LDA and BLR can be used to effectively classify patients with severe and non-severe COVID-19, even with violation of the normality assumption.

Covid-19 Diagnosis: Comparative Approach Between Chest X-Ray and Blood Test Data

The Covid-19 virus has made a major impact on the world and is still spreading rapidly. A reliable solution to prevent further damage, early diagnosis of coronavirus patients are incredibly important. While chest X-Ray diagnosis is the easiest and fastest solution for this, an average radiologist has only a 75% to 85% accuracy when evaluating X-Ray data, thus it is desirable to achieve an accurate artificial network for this. Throughout this study, chest X-Ray data and blood routine test data are utilised and compared. X-Ray data consists of 5000 chest X-Ray images which are gathered from an open-source research and from a local hospital in which both have anonymous data. The blood test results were also taken from the same hospital. For the chest X-Ray diagnosis we utilised two of the popular convolutional neural networks, which are Resnet18 and Squeezenet and concluded that Resnet18 provided slightly more accurate results, while both having almost 98% accuracy. For blood test diagnosis, a feed-forward multi layer neural network was used. Even though it was worked on an insufficient dataset, 72% accuracy was obtained, thus making it a feasible option for further research. Hence, we concluded that in general chest X-Ray diagnosis is preferable over routine blood test diagnosis and the usage of AI yields better approximate results than humans. © 2021 IEEE

Smart Surveillance for COVID-19 Health Protocol Compliance in Vehicles

Nowadays, the Corona Virus outbreak in 2019 (COVID-19) has become a global pandemic. The public must implement health protocols to reduce the spread of COVID-19. Trends show that the number of COVID-19 is increasing over time. This study proposes and develops a smart model to detect COVID-19 Health protocol violators in vehicles. This model can detect violations of the use of masks and social distancing in vehicles. The proposed model is a combination of the YOLO object detection method and the Hourglass architecture. The experimental results of the proposed model can detect violations with a high success rate. Here, the standard YOLOv4 detection model as baseline yields an mAP of 0.87 for validation and 0.74 for test data. On the other hand, the proposed method produces an mAP of 0.92 on the validation data, 0.78 on the test data. From these results, this smart model is quite promising to help reduce the spread of COVID-19. © 2021 ACM.

Diagnosis of COVID-19 from chest X-ray images using Deep Learning Model

COVID-19 (Coronavirus-2019) is a disastrous pandemic which has affected the whole world damaging the whole ecosystem specially health. Researchers around the world have been contributing to the advancement in these conditions around the world. Medical practitioners have found that chest X-Ray images can used to detect whether a person suffers from covid-19 or not due to anomalies in chest radiography images. Continuing this motivation we have taken around 700 chest X ray images from different resources and applied two different transfer learning techniques to build a model which can detect the existence of covid-19 in a person. It uses VGG16 and Resnet50 deep learning models which utilize transfer learning to train their parameters. We have trained our both models for 50 epochs. VGG16 gives 76% of accuracy on test data and Resnet50 gives 85% accuracy on test data. We have tried to engineer thresholds for probability of classification thus changing specificity and sensitivity and also evaluated our models on various metrics such as classification report, confusion matrix heatmap, roc-auc score. This is by no means to be used for any medical procedures but can help other researchers to take some useful insights from it and carry forward the learning in building something which is production ready for contribution in our fight against COVID-19. © 2021 IEEE.

FaceNet with RetinaFace to Identify Masked Face

The use of masks due to the Covid-19 pandemic reduces the accuracy of facial recognition systems applied to camera-based security systems. The use of the mask by the people covers most of the facial featureswhich is located from middle to bottom area. In addition, the area which are still visible are the upper face which are eyes and forehead. This paper proposes a masked face recognition using a combination of RetinaFace as a face detector and FaceNet as a face recognizer. The MFR2 dataset with 53 identities was used to train and test this method. The test data in this study are only images of masked faces. Cosine Distance was implemented to measure the face similarity. Based on the experiment results, the proposed method obtained 98.2% of detection accuracy. The proposed method provided 78% accurate performance with 3.63 s for processing a single frame in terms of face recognition. The performance indicates that our system can potentially be applied in security systems with many different identities. © 2021 IEEE.

Work in Progress: Leveraging a Virtual Precollege Summer Coding Day Camp to Promote DEI in Recruiting Students to Computer Science and Information Technology

As part of an ongoing initiative to recruit students to the Computer Science and Information Technology degree programs at Southeastern Louisiana University, a summer coding day camp was formed beginning in the summer of 2019 through a grant with Louisiana Economic Development Fast Start. The 2019 camp was a two-week on-campus experience. In the success of the first year, expansion, to include a satellite campus, was planned for year two. This was never realized due to COVID-19. The summer 2020 delivery and curriculum was redesigned two short months before delivery. The decision was made to offer a much abbreviated online version of the camp, while maintaining the maximum capacity. Through a partnership with cyber.org, curriculum was selected and a virtual capture-the-flag was offered. The capture-the-flag competition served to promote participation in the recruitment activities. Through the use of pre and post tests, data was collected as to familiarity with the university, the Department of Computer Science degree offerings, job opportunities in the field, and intention to attend college. Additionally, student surveys were administered to collect demographic information. This paper details the experience of offering a virtual summer coding camp and explores both the challenges and opportunities that were encountered. Details into the specifics of how the camp was administered and recruiting activities are presented as are the results of the survey findings. It is concluded that the experience was a success, reaching maximum enrollment within 48 hours and achieving a wait-list of over 80. Of the students enrolled in the camp, women and minorities represented 50% of the students and the 80% of the students reported that their expectations were met or exceeded. © American Society for Engineering Education, 2021

Risk factors for positive and negative COVID-19 tests: a cautious and in-depth analysis of UK biobank data.

BACKGROUND: The recent COVID-19 outbreak has generated an unprecedented public health crisis, with millions of infections and hundreds of thousands of deaths worldwide. Using hospital-based or mortality data, several COVID-19 risk factors have been identified, but these may be confounded or biased. METHODS: Using SARS-CoV-2 infection test data (n = 4509 tests; 1325 positive) from Public Health England, linked to the UK Biobank study, we explored the contribution of demographic, social, health risk, medical and environmental factors to COVID-19 risk. We used multivariable and penalized logistic regression models for the risk of (i) being tested, (ii) testing positive/negative in the study population and, adopting a test negative design, (iii) the risk of testing positive within the tested population. RESULTS: In the fully adjusted model, variables independently associated with the risk of being tested for COVID-19 with odds ratio >1.05 were: male sex; Black ethnicity; social disadvantage (as measured by education, housing and income); occupation (healthcare worker, retired, unemployed); ever smoker; severely obese; comorbidities; and greater exposure to particulate matter (PM) 2.5 absorbance. Of these, only male sex, non-White ethnicity and lower educational attainment, and none of the comorbidities or health risk factors, were associated with testing positive among tested individuals. CONCLUSIONS: We adopted a careful and exhaustive approach within a large population-based cohort, which enabled us to triangulate evidence linking male sex, lower educational attainment and non-White ethnicity with the risk of COVID-19. The elucidation of the joint and independent effects of these factors is a high-priority area for further research to inform on the natural history of COVID-19.