Comparación de la Discretización Estándar con un Nuevo Método para Reglas de Asociación Cuantitativas
(Comparison of Standard Discretization with a New Method for Quantitative Association Rules)
Juan Luis Domínguez Olmedo (firstname.lastname@example.org)1, Jacinto Mata Vázquez (email@example.com)1
1Universidad de Huelva
This paper appears in: Revista IEEE América Latina
Publication Date: April 2016
Volume: 14, Issue: 4
In the task of extracting association rules in datasets with numerical attributes, the usual approach is to previously discretize those attributes. In this paper, a deterministic method for the extraction of quantitative association rules is presented, which does not employ a previous discretization. It also tries to eliminate redundant rules and reduce the search. Experiments have been performed comparing it with the well-known deterministic algorithm Apriori, and statistical validation of the results was carried out using nonparametric tests. From the results obtained, the proposed approach can be seen as a suitable deterministic way of extracting quantitative association rules without a previous discretization.
data mining, association rules, discretization
Documents that cite this
This function is not implemented yet.
[PDF Full-Text (338)]