سميرا طالبي اناركي

Clustering of Covid-19 data using different clustering algorithms and also predicting its spread pattern has become a global topic. The high preva‎lence of covid-19 has made it a new epidemic. Forecasting the worldwide preva‎lence and incidence is important to help health professionals make key decisions. In this study, our aim is to cluster and predict the incidence of covid-19 in ten countries of the world for better management of this disease. On these data of death and infection due to covid-19, k-means data mining algorithms for clustering and LSTM algorithm have been implemented to predict the preva‎lence pattern for each country. The rapid spread of the Covid-19 disease has become a serious threat to the world. So far, millions of people around the world have been infected with this disease. The Covid-19 pandemic has had significant effects on various aspects of human life. In order to be safe and make necessary arrangements, it is necessary to predict the spread of the virus at this time. This can help control the rate of epidemics and deaths. Previous studies have mainly used statistical tools and algorithms based on machine learning. However, the former were inadequate for the unpredictable analysis of the epidemic, and the latter mainly experienced problems of underfitting or overfitting. To overcome these problems, the present study has proposed a method based on deep learning on long-term data. In this thesis, an approach for predicting cases of infection and death due to covid-19 based on LSTM long-short-term memory networks was presented. The LSTM model was implemented on the time series data of 10 European countries in the mentioned period and the RMSE eva‎luation criterion was calculated. The results showed that the LSTM neural network had a good performance in predicting cases and deaths in ten countries: Belgium, Finland, France, Germany, Greece, Italy, Romania, Poland, Spain and Sweden. Estimating the infected and dead people due to Covid-19 can help control the pandemic situation.

كوويد-19 , الگوريتم خوشه‌بندي k-means , الگوريتم LSTM , پيش‌بيني , سري زماني

Covid-19 , K-Means Clustering algorithm , LSTM algorithm , prediction , time series

Samira Talebi Anaraki

Javad Vahidi