Estrutura para Utilização de Recuperação de Imagens Baseada em Conteúdo em Oráculos de Teste de Software com Saída Gráfica

Decision-Rule Solutions for Data Mining with Missing Values

Rafael A. P. OliveiraMárcio E. DelamaroFátima L. S. Nunes

Sholom M. WeissNitin Indurkhya

This paper presents a prototype of a system whose goal is to highlight the opportunity to explore computer vision applied in the Content-based Image Retrievial (CBIR), in order to testing oracles for software that generate graphical output. Using libraries of Java programming language Java, a structure that allows the alliance of two such disparate areas in the current scenario of computing has been implemented. A great flexibility was given to the structure, so that the user chooses the way of analysis of images. In this context it is possible such a user choose the characteristics that must be extracted from the images and how these should be considered in the test, thus creating what was called 'Graph Oracle'. A method is presented to induce decision rules from data with missing values where (a) the format of the rules is no different than rules for data without missing values and (b) no special features are specified to prepare the the original data or to apply the induced rules. This method generates compact Disjunctive Normal Form (DNF) rules. Each class has an equal number of unweighted rules. A new example is classified by applying all rules and assigning the example to the class with the most satisfied rules. Disjuncts in rules are naturally overlapping. When combined with voted solutions, the inherent redundancy is enhanced. We provide experimental evidence that this transparent approach to classification can yield strong results for data mining with missing values.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: