Resumo

ROJAS DELGADO, Jairo. Feature selection on pap-smear data using heuristic information. Rev cuba cienc informat [online]. 2016, vol.10, n.2, pp.73-88. ISSN 2227-1899.

This paper presents a method of feature selection to classify cell samples taken from the cervix into cancerous or not cancerous cells. Cancer of the cervix is the second most widespread cancer in women and the investigation of computational methods based on the correct classification of samples has a direct impact on the life's quality of patients. A set of selected features is investigated taking into account the importance of these features and the ability of a test group of classifiers to assess those traits. The test classifiers used are: HCM, FCM, LS and WKNN. The results demonstrate that the use of that set of features generally yields good classification results.

Palavras-chave : classification; discriminator potential; feature selection; pap-smear.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )