Tamanho da fonte: 

The search for new knowledges is one of the human characteristics. New knowledges can be obtained from a set of already existent knowledges. This is the case of the KDD - Knowledge Discovery in Database, which intends to provide the capacity of discovering new and significant information for the use of existent data. This paper, following the approach of the Waterfall Data Input Process, proposes a method of weight inference of the value to be attributed. This new method takes into account measures of validation of the groups produced by the process of grouping (clustering), the use of quality concepts for cluster classification and measures of distance of the neighbors.

KDD; Data Imputation; k-Means; k-NN; Silhouette