Pre-Diagnosis of Chronic Diseases by the Application of Computational Intelligence Models

Abstract

Computational intelligence models applied to medicine have become a growing area of research around the world. This article presents the results of a documentary research that allows identifying the most important computational intelligence models of the state of the art that are applied in the pre-diagnosis of diseases, as well as a study of the algorithms that represent each of these models. Also, through the use of the WEKA platform and the KEEL and UCI repositories, the performance exhibited by these pattern classifiers when applied in the pre-diagnosis of some important chronic diseases in the context of human health is studied. In particular, algorithms representing the most appreciated approaches in the area of intelligent pattern classification are tested. In the experimental part of this article, these computational intelligence algorithms were applied in databases of breast cancer, hyperthyroidism, and hypothyroidism. The results that were derived from the experiments show the superiority of the performance of the vector support machines, the 1-NN classifier and the Lernmatrix tau[9]. The performances achieved allow confirming, with a high degree of certainty, the usefulness of computational intelligence models in the pre-diagnosis of chronic diseases.