Minería de datos para la identificación de similitud en la información de pacientes

Abstract

In the present work order degree, it aims to identify groups of similar patients based on disease. Guided by the CRISP-DM methodology, some experiments were done using clustering methods as BDSCAN, and K-means clustering and semantic methods as LSA. Once the different groups was obtained, could be identified, the most common diseases in each group. Individual and group analysis result of experimentation was performed. Finally validation of all patient groups was performed found