Your browser doesn't support javascript.
COVID19 Prediction using Time Series Analysis
Proc. - Int. Conf. Artif. Intell. Smart Syst., ICAIS ; : 1599-1606, 2021.
Article in English | Scopus | ID: covidwho-1219398
ABSTRACT
The ongoing COVID19 pandemic has created havoc all over the world. Millions of lives have been gone and thousands are vulnerable. It has also affected the world economy due to lockdown. So, there is a need to develop a time-series forecasting model for predicting future cases so that necessary precautions can be taken. The aim is to help in coping up with the situation without affecting lifestyle any further. For forecasting, this paper has considered four models exponential smoothing, ARIMA and SARIMA models, and identifies the suitability of the model for prediction purposes. Also, to incorporate the impact of festive seasons in India to measure the fluctuations in the new cases, SARIMAX model is used. The spread case data of the pandemic collected is of 47 weeks, from 30th January 2020 to 23 r d December 2020 for India. Data of the first 45 weeks (90%) is taken for training the model and that of the last 2 weeks (10%) is used for validation purposes. For evaluation purposes RMSE (root mean square error) and MAE (mean absolute error) are taken as parameters for model evaluation. The ARIMA (8,2,1) model performed best among all models based on the RMSE. Considering multiplicative trend and seasonality Triple Exponential Smoothing model gave the best result with respect to MAE. © 2021 IEEE.

Full text: Available Collection: Databases of international organizations Database: Scopus Type of study: Experimental Studies / Prognostic study Language: English Journal: Proc. - Int. Conf. Artif. Intell. Smart Syst., ICAIS Year: 2021 Document Type: Article

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: Databases of international organizations Database: Scopus Type of study: Experimental Studies / Prognostic study Language: English Journal: Proc. - Int. Conf. Artif. Intell. Smart Syst., ICAIS Year: 2021 Document Type: Article