Machine learning predicting COVID-19 in Algeria

Predicting new and urgent trends in epidemiological data is an important public health problem. This problem has gained increasing attention of the data mining and machine learning research communities. Artificial Intelligence can extract relevant information from an increasingly accessible dataset that would be difficult to navigate in a nonautomated way. In this paper, our goal is to propose a new spatiotemporal system for predicting COVID-19 cases in the 48 cities of Algeria. This system is mainly based on AI methods, namely: ARIMA, LSTM, SLSTM and Prophet. Real-time data collection was used in our study. The dataset is randomly split into training set and testing set. In our prediction experiment, comparison between observed and predicted values using performance metrics were performed and obtained results were very satisfactory and stable and responses fit completely to each other. The main purpose of our system is to help public health officials and planners to manage services and organize medical infrastructure as well as evaluate action plans to deal with future course of the COVID-19 epidemic.