Selecting effective action strategies for the participants in a hospitalization process with the use of a fuzzy cooperative game and a genetic algorithm

Abstract

Introduction: The use of linear programming methods in making decisions on hospitalization in a fragile epidemiological situation may be hampered by the necessity to take account of a large number of parameters and limitations of the participants. Purpose: Development of an approach to selecting effective action strategies for the participants in a hospitalization process, with social factors taken into consideration. The approach is based on the theory of cooperative games which are solved with the use of a genetic algorithm. Results: A cost function has been developed for evaluating the effectiveness of the hospitalization process on the basis of the selected strategies and in consideration of social factors. A genetic algorithm has been designed in which the proposed effectiveness evaluation function is used as a fitness function for a population, while to determine chromosomes of individuals in the population the set of selected strategies of the hospitalization process participants is used. The approach has been tested using the data on hospitalizations of patients with suspected COVID-19, that were provided by several ambulance stations in Saint-Petersburg, Russia. The study shows the superiority of the proposed approach over the previously developed one in terms of the speed of solving a cooperative game, the quality of the solution being maintained. Practical relevance: Some software which is based on the proposed approach can be integrated into an ambulance dispatcher’s automated workstation to support decision-making during the process of hospitalization in a fragile epidemiological situation.