Feature selection for the classification of clinical data of stroke patients

Setyawati, Onny and Arifianto, Aji Seto and Sarosa, Moechammad (2017) Feature selection for the classification of clinical data of stroke patients. In: 2017 20th International Conference on Electrical Machines and Systems (ICEMS). IEEE, Sydney, NSW, Australia, pp. 1-4. ISBN 978-1-5386-3246-8

	Text (Full Text) setyawati2017.pdf - Published Version Download (327kB)
	Text (Hasil Similarity) 5. Cek Sim_Feature Selection.pdf - Supplemental Material Available under License Creative Commons Attribution Share Alike. Download (1MB)

Abstract

Clinical examination of the patient with suspected stroke to determine the type of pathology is still widely applied, especially in Indonesia due to constraints in the implementation of the Gold Standard Procedure. Clinically, the examination of the various features starts from the physical symptoms, medical history and laboratory results, which might take long duration and costly. Moreover, not all inspection features have a significant influence to distinguish the type of stroke, hence, sorting features are required. The selection process to get the best features is performed by identifying similarity to the features of each class. Fuzzy Entropy generates the entropy value from the degree of membership of each feature. The result of the implementation of feature selection is able to select 13 of the best features with 96% in accuracy, therefore, the process is more effective than having to check 32 features.

Actions (login required)

View Item