Development of forecasting model for infectious disease (COVID-19)

Abstract

The spread of COVID-19 in the whole world has placed mankind in harm's way. The resources of unquestionably the greatest economies are concerned due to the immense infectivity and infectiousness of this ailment. The limit of ML models to measure the amount of impending patients impacted by infectious disease which is currently considered as an imaginable threat to mankind. It was first perceived in December 2019 in Wuhan, the capital of China's Hubei area. The objective of this investigation is to propose an envisioning model using the COVID-19 open dataset from top impacted regions across the world using AI computations. Simulated intelligence figurines help us with achieving this objective. Backslide models are one of the controlled AI strategies to aggregate tremendous degree data. This investigation intends to apply Multivariate Linear Regression to predict the amount of asserted and destroyed COVID-19 cases for a scope of one and fourteen days. The test outcomes explain 90% irregularity in conjecture. The computations are surveyed using the screw up organization, for instance, Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and precision for top affected locales across the world. © 2021 American Institute of Physics Inc.. All rights reserved.