Forecasting By Time Series Models For Covid-19 Cases In India – A Comparative Study

Abstract

The objective of the study is to get a overall understanding of the spread of covid-19 in India overtime. The study has attempted to find the trend of the confirmed, cured and death cases are visualized and also compared for different states in India. Time series models like ARIMA, Exponential, Artificial Neural Network, Moving average method, Root mean square method is used. Finally ARIMA, Exponential and Artificial Neural Network is compared to find which method is best. Based on Root Mean Square value the best fitted model for forecasting future values are selected. Thus a fitted Time Series model is used to forecast a trend line for next 6 months.