Systematic Literature Review: Machine Learning Prediction Model for Covid-19 Spreading

ABSTRACT

The dataset, methods, and machine learning prediction framework on the Covid-19 theme have been published widely and complex. Special publications on the spread of virus infection 19 in the form of a time series need to be mapped more comprehensively. This literature review aims to identify and analyze research trends, datasets, and methods used in predicting Covid-19 with Machine Learning Engineering research between 2019 and 2021. Identifying the need, specifying the research question evaluating review protocol, searching for papers, scanning papers, and reporting results are the eight major steps of this systematic literature review. The most critical aspect of systematic analysis is defining the research questions. The PICOC techniques are used to identify research questions. Journal candidates were filtered out using inclusion and exclusion criteria techniques to shrink the SLR scope area. based on a literature study it was found that research in 2019-2021 on the Covid-19 distribution prediction system used variables susceptibility, infection, mortality, geography, weather, and patient clinical data to be processed into ANFIS machine learning prediction models and neural networks are several models. A classification model that is widely used for hybrid processing in calculating covid-19 infection prediction. The datasets that are often used do not fully meet the epidemiological aspects that trigger the spread of COVID-19 infections. ANFIS and NN are several classification methods that are widely used for hybrid processing in calculating predictions of the spread of COVID-19 infection. © 2022 IEEE.