Adjustment of an Epidemiological Cellular Automata-based Model using Genetic Algorithm

Reliable modeling allows the simulation of critical processes that can serve as a foundation for planning and defining public policies. Ecological, climatic, public health and epidemiological models, among others are important research instruments that can forecast and evaluate the impact of decisions made by organizations and governments. Once the basic representation of the process is defined, one of the main difficulties of modeling is the adjustment of several parameters that make up it. We investigate the application of genetic algorithms to adjust model parameters relying on data series as input since they consist in a powerful adaptive search method. The proposed approach is evaluated using a previous model based on probabilistic cellular automata that describes the evolution of a population of insect vectors responsible for Chagas disease. The experiments performed here shown that results of the evolutionary parameters adjustment are similar to the behavior of the reference model both in the quantity of insects and in their spatial distribution. Our approach achieved a robust error of 3.13, that is, a difference of approximately 3 insects in one-year simulation.