Ricardo B. C. Prudêncio,  Teresa B. Ludermir,  Francisco de A. T. de Carvalho. 

The selection of a good model for forecasting a time series is a task that involves experience and knowledge. A promising approach to acquire knowledge for this task is the use of machine learning algorithms. In this work, we proposed the use of a novel learning algorithm, the Meta-Prototypes (MP), for the model selection problem. This algorithm is related to Symbolic Data Analysis, which is a new area in the field of knowledge discovery. In our work, the MP algorithm was used in a case study and it was compared to some traditional learning algorithms. So far, the MP algorithm obtained the lowest selection error among all the tested algorithms. Clique no link abaixo para buscar o texto completo deste trabalho na Web: Buscar na Web Processando consulta.