An evolutionary two-objective genetic algorithm for asthma prediction

Genetic Algorithms in combination with Artificial Neural Networks have been used to solve optimization problems in several domains. In this paper, an evolutionary algorithm consisting of an Artificial Neural Network and a Genetic Algorithm is presented for predicting the asthma outcome in children under the age of five. The most cases of asthma begin during the first years of life, thus the early determination of which young children will have asthma later in their life counts as an important priority. A Genetic algorithm search is implemented in order to investigate which prognostic factors contribute most to the asthma prediction. This search results to pruned input and hidden layers of the Artificial Neural Network as well as minimization of the Mean Square Error of the trained network at the test phase. Thus, dimension reduction of the prognostic factors can be achieved without any loss of prediction ability.