COVID-19 epidemic: analysis and prediction

Abstract

“Novel Coronavirus”, commonly known as COVID-19 has spread nearly to the entire world. The number of impacted cases and deaths has increased significantly in each country, posing a challenge for the world’s health organizations. The goal of this paper was to better comprehend and analyze the growth of the disease in India, including confirmed, recovered, fatalities, and active cases of COVID-19. Data analysis affects an organization’s decision-making process with interactive visual representation. The proposed model was an ensemble model that was built using linear regression, polynomial regression, and support vector machine (SVM) regression models. The model predicted the number of confirmed cases from 30 th May 2021 to 15 th June 2021 based on the data available from 22 January 2020 to 29 May 2021 and improved accuracy was obtained when compared with the actual data. Forecasting the confirmed cases might assist health organizations in planning medical facilities. Following that, an appropriate machine leraning (ML) model must be found that can predict the number of new cases in the future.

A brief review and scientometric analysis on ensemble learning methods for handling COVID-19

“…Using a shark optimization algorithm and improved bat algorithm with multi-class SVM A collection of CT chest images uploaded to GitHub, including 349 images from 216 patients in several hospitals in China 8 Automated diagnosis of childhood pneumonia in chest radiographs using modified densely residual bottleneck-layer features [ 72 ] Highest accuracy: 99.6% Adaboost 5232 CXR images for children aged one to five years old from Kaggle, of which 3883 were infected and 1349 were normal 9 Multi-channel transfer learning of chest x-ray images for screening of COVID-19 [ 73 ] Highest accuracy: 94% Recall: 100% Three ResNet-based models for one-to-all classification Chest X-ray images including 1579 normal, 4245 pneumonia and 184 COVID-19 cases 10 Ensemble learning for poor prognosis predictions: A case study on SARS-CoV-2 [ 74 ] Highest accuracy: 95% Seven prediction models defined in China named Dong, Shi, Gong, Lu, Yan, Xie, and Levy 5394 cases in two hospitals in China, London King's College Hospital and University Hospitals Birmingham 11 Using Automated Machine Learning to Predict the Mortality of Patients With COVID-19: Prediction Model Development Study [ 75 ] Highest accuracy for stacking model: 79.1% Using a real-time method and building 20 different machine learning models through auto ML. Model interpretation through Shapley's additive explanation and dependency graphs for extracting 10 influential variables - using a binary classifier 4313 cases in Albert Einstein College of Pharmaceutical Sciences in New York 12 COVID-19 epidemic: analysis and prediction [ 76 ] Highest accuracy: 99.94% Using linear regression, polynomial regression and SVM 14,654 Indian patients' data from Johns Hopkins University Center for Science and Engineering 13 Deep ensemble model for classification of novel coronavirus in chest X-ray images [ 77 ] Deep CNN model namely MobileNet, ResNet50 and InceptionV3. …”

Impulse noise recuperation from grayscale and medical images using supervised curve fitting linear regression and mean filter

“…An S-curve logistic regression technique is mastered in [18] for the prediction of diabetes. The linear regression technique [19], [20] is applied for the automatic prediction of missing values. A threshold-based technique is employed to reduce noise, and a locally adaptive threshold-based edge preservation denoising scheme is used [21].…”