ABSTRACT
Machine learning has seen a considerable increase in performance and interest in scientific research and industrial applications over the previous decade. The success of most current state-of-the-art methods can be linked to recent deep learning advancements. Deep learning has been demonstrated to outperform not only standard machine learning but also highly specialized tools designed by domain specialists when applied to many scientific fields involving the processing of non-tabular data, such as pictures or text. This article will cover ML-based research on SARS-Co V-2 Proteinase Biological Activity classification, with an emphasis on the most recent successes and research trends. SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus 2) produced a global pandemic of coronavirus illness (COVID-19), which prompted a rush to find treatment options. Despite the attempts, no vaccine or medicine for therapy has been approved. In this paper, we mentioned some previous articles that have resulted in successful bioactivity prediction. The discussion of the machine having to learn technology that has been used for bioactivity prediction in general and has the potential to lead the way for successful working with complex molecules in the future is also a focus of this review. The study finishes with a brief viewpoint on contemporary machine learning research advances, including student engagement and semi-supervised learning, which offer considerable potential for increasing bioactive discovery. © 2022 IEEE.
ABSTRACT
In this current COVID-19 scenario, an effective face mask detection application. The project's major purpose is to put this system in place at college entrances, airlines, hospitals, and offices where the risk of COVID-19 spreading through contagion is highest. According to reports, having a face mask while at work significantly minimizes the chance of transmission. It's an issue of object detection and classification with two classes (Mask and Without Mask). For recognizing face masks, a hybrid model combining deep and traditional machine learning will be shown. This face mask detector is built with Python, OpenCV, TensorFlow, and Keras and is based on a dataset. Everyone should inspect their face before entering the building and make sure they have a mask with them. A beep alert will be triggered if somebody is found without a face mask. As a result, all of the workplaces are reopening, the number of instances of COVID-19 being reported around the country is steadily rising. It can be brought to a close if everyone observes the safety precautions. As a result, we expect that this research will assist in detecting people wearing masks to work. © 2021 IEEE.
ABSTRACT
Nowadays, Covid-19 is one of the major problems in the world. It is spread very quickly by connecting or touching with a covid positive person. To detect the covid-19, we have to use the testing kits. But we don't have that many kits for testing the covid-19 because the affected number of people is increasing day by day. To solve these big issues, we are introducing one another method. To detect the covid-19 we need to use either chest X-ray's image or Computed Tomography (CT) images. The reason behind to implement of the model is very simple and easy because almost every hospital diagnostic center has X-rays imaging facilities. To identify the covid positive or negative cases, we do not require any kits. In this article, we are introducing one novel model, the process of building the model, and the dataset that we have used to train our model. To train the model we have used almost 1000 chest X-ray images and 700 CT images. For training the model, we are using deep learning algorithms like VGG16, VGG19, Inception V3, RestnetSO, and Xception. We also compare all of the algorithms with some comparison graphs. Among all of the deep learning models, the Inception V3 performs the best accuracy in both datasets. © 2021 IEEE.