Pattern analysis: predicting COVID-19 pandemic in India using AutoML

Abstract

Purpose Since December 2019, global attention has been drawn to the rapid spread of COVID-19. Corona was discovered in India on 30 January 2020. To date, in India, 178,014 disease cases were reported with 14,011 deaths by the Indian Government. In the meantime, with an increasing spread speed, the COVID-19 epidemic occurred in other countries. The survival rate for COVID-19 patients who suffer from a critical illness is efficiently and precisely predicted as more fatal cases can be affected in advanced cases. However, over 400 laboratories and clinically relevant survival rates of all present critically ill COVID-19 patients are estimated manually. The manual diagnosis inevitably results in high misdiagnosis and missed diagnosis owing to a lack of experience and prior knowledge. The chapter presents an option for developing a machine-based prognostic model that exactly predicts the survival of individual severe patients with clinical data from different sources such as Kaggle data.gov and World Health Organization with greater than 95% accuracy. The data set and attributes are shown in detail. The reasonableness of such a mere three elements may depend, respectively, on their representativeness in the indices of tissue injury, immunity and inflammation. The purpose of this paper is to provide detailed study from the diagnostic aspect of COVID-19, the work updates the cost-effective and prompt criticality classification and prediction of survival before the targeted intervention and diagnosis, in particular the triage of the vast COVID-19 explosive epidemic. Design/methodology/approach Automated machine learning (ML) provides resources and platforms to render ML available to non-ML experts, to boost efficiency in ML and to accelerate research in machine learning. H2O AutoML is used to generate the results (Dulhare et al., 2020). ML has achieved major milestones in recent years, and it is on which an increasing range of disciplines depend. But this performance is crucially dependent on specialists in human ML to perform the following tasks: preprocess the info and clean it; choose and create the appropriate apps; choose a family that fits the pattern; optimize hyperparameters for layout; and models of computer learning post processes. Review of the findings collected is important. Findings These days, the concept of automated ML techniques is being used in every field and domain, for example, in the stock market, education institutions, medical field, etc. ML tools play an important role in harnessing the massive amount of data. In this paper, the data set relatively holds a huge amount of data, and appropriate analysis and prediction are necessary to track as the numbers of COVID cases are increasing day by day. This prediction of COVID-19 will be able to track the cases particularly in India and might help researchers in the future to develop vaccines. Researchers across the world are testing different medications to cure COVID; however, it is still being tested in various labs. This paper highlights and deploys the concept of AutoML to analyze the data and to find the best algorithm to predict the disease. Appropriate tables, figures and explanations are provided. Originality/value As the difficulty of such activities frequently goes beyond non-ML-experts, the exponential growth of ML implementations has generated a market for off-the-shelf ML solutions that can be used quickly and without experience. We name the resulting work field which is oriented toward the radical automation of AutoML machine learning. The third class is that of the individuals who have illnesses such as diabetes, high BP, asthma, malignant growth, cardiovascular sickness and so forth. As their safe frameworks have been undermined effectively because of a common ailment, these individuals become obvious objectives. Diseases experienced by the third classification of individuals can be lethal (Shinde et al., 2020). Examining information is fundamental in having the option to comprehend the spread and treatment adequacy. The world needs a lot more individuals investigating the information. The understanding from worldwide data on the spread of the infection and its conduct will be key in limiting the harm. The main contributions of this study are as follows: predicting COVID-19 pandemic in India using AutoML; analyzing the data set predicting the patterns of the virus; and comparative analysis of predictive algorithms. The organization of the paper is as follows, Sections I and II describe the introduction and the related work in the field of analyzing the COVID pandemic. Section III describes the workflow/framework for AutoML using the components with respect to the data set used to analyze the patterns of COVID-19 patients.

Detail-Oriented Capsule Network for classification of CT scan images performing the detection of COVID-19

“…Gomathi et al [23] offer an option to develop a predictive model on the machine, which predicts the continuation of severe patients, such as Kaggle dat.gov and the World Health Organisation, with more than 95 percent accuracy. The rationality of only three elements can depend on the tissue injury, immunity and inflammation indicators.…”

Designing optimal COVID-19 testing stations locally: A discrete event simulation model applied on a university campus

“…In accordance with the Centers for Disease Control and Prevention (CDC), proactive testing for COVID-19 infection is a key factor in determining where and how the SARS-CoV-2 virus is spreading within a population. The early identification of infected people leads to more rapid treatment and isolation for them, as well as for those who were exposed to them [1][2][3]. This type of monitoring is essential to reduce the spread of the disease (CDC, 2020).…”

“…test the generalizability of this COVID-19 testing process in other situations. A promising line for future research would be to combine such simulation models with newly developed artificial intelligence techniques, e.g., automated machine learning [3], deep learning techniques [47] to further predicting and mitigating the COVID-19, as well as to share and maintain these data in a transparent and decentralized way using Blockchain technology [48].…”

Early detection of COVID-19 outbreaks using human mobility data

“…To appropriately target the timing, location, and severity of such measures, it is essential to be able to predict when and where outbreaks will occur and how widespread they will be. Recent work has focused on using past health data to forecast accumulated COVID-19 cases and deaths using exponential smoothing models [10,11], autoregressive moving average models [11][12][13], and deep learning models [13][14][15][16][17][18].…”