Weighted Fuzzy Time Series Model to Forecast Epidemic Injuries and its Data Visualization

Coronavirus disease spread widely around the world. We have to make accurate future forecasts to predict the amount of COVID-19 infections to help make decisions. In this paper, we present some of the techniques that have proven useful in predicting the number of people infected with COVID-19 in a specific period in different regions of the world on the basis of gaining global experience from which we are working to improve our future methodologies to face such a pandemic and in an effort to discover the optimal application.

In order to forecast injuries in regions with COVID-19, particularly in the Arab region, we adopt a rewarding model. This forecast uses epidemic injuries data from March 2nd, 2020 to July 20th, 2020 in Saudi Arabia. Time series data is displayed in two different ways: charts and graph visualization.

To compare with the traditional Auto-Regressive Integrated Moving Average (ARIMA) statistical method, we suggest using weighted fuzzy time series methods (WFTS) and weighted non-stationary fuzzy time series techniques (WNSFTS). The provided data is not stationary, hence stationary data conversion is necessary before using (ARIMA) and (WFTS) algorithms to forecast it. We do a log transform and differencing on our injuries dataset.

The visual Graph Describes Covid Situation in Saudi Arabia as Improving. When we use the Dickey-Fuller Test (DFT) to analyse the original data, we discover that the p-value is equal to 0.646, which is greater than 0.05 and suggests non-stationarity. To compare the accuracy of the approaches, the mean square error (MSE), root mean square error (RMSE), and normalization root mean square error (NRMSE) are used. The findings demonstrate that WFTS methods provide good services for estimating COVID-19-based epidemic injuries in the region.

When projecting the problem of epidemic injuries, the application of Weighted Non Stationary Fuzzy Time Series (WNSFTS) can lead to noticeably better outcomes. Because it has a high level of predictive accuracy and can foretell future cases of infection.