ANALYSIS OF INFECTIOUS DISEASES FORECASTING METHODS

Abstract

At the end of 2019, COVID showed the world its unpreparedness and inability to resist the modification of the influenza virus. The onset of a global pandemic, the spread of the disease, and the large number of deaths have led not only to the search for control of the virus, but also to the possibility of predicting its spread. While some scientists have developed a vaccine, others have studied the prospects of the virus, filling the planet and predicting the number of deaths under certain conditions. Using statistical data, the researchers developed maps of the spread of the virus, possible future targets, and even estimated possible deaths from various strains of COVID-19.

The main task of data forecasting is to create some models from the provided data set in order to provide useful and correct forecasting of future or unknown values of another data set.

For many years, standard statistical methods and mainly the intuition of the doctor, his knowledge and experience have been used to predict the risk of the disease, the occurrence of complications in the patient, the spread of the disease among other people. This approach to disease assessment often leads to unwanted biases, errors and large losses. In the modern field of medicine, such an assessment also has a negative impact on the quality of services provided to patients. With the availability of electronic health data, more reliable and advanced computational approaches have emerged, such as machine learning and big data analysis. The emergence of new methods in data mining has led to the study and application of forecasting methods in the field of disease. In the literature, most relevant studies have used one or more machine learning algorithms to predict a particular disease. For this reason, comparing the effectiveness of different controlled machine learning algorithms for disease prediction is the main focus of this study.