Diagnosis of diabetes mellitus using PSO and KNN classifier

摘要

Diabetes is a complex disease whose prevalence is constantly increasing due to lifestyle changes and an aging population. This paper presents a pre-processing step "the selection of attributes” that plays an important role in data mining. It allows to build a model describing the data by removing the redundant, irrelevant or noisy attributes. Applied to the task of classification in data mining, it ensures a reduction in the size of the problem, which reduces the duration of learning and simplifies the learned model. This simplification generally facilitates the interpretation of this model. It also makes it possible to avoid the phenomenon of over-learning improving the accuracy of the prediction and the understanding of the classifier. In this approach the KNN classifier is used for classification.