RT Journal Article
T1 Redundancy Is Not Necessarily Detrimental in Classification Problems
A1 Grillo, Sebastián Alberto
A1 Vázquez Noguera, José Luis
A1 Mello Román, Julio César
A1 García Torres, Miguel
A1 Facon, Jacques
A1 Pinto Roa, Diego Pedro
A1 Salgueiro, Luis Fernando
A1 Bareiro Paniagua, Laura Raquel
A1 Leguizamón Correa, Deysi Natalia
A2 Universidad Americana (PY)
A2 Universidad Nacional de Concepción (PY)
AB In feature selection, redundancy is one of the major concerns since the removal of redun dancy in data is connected with dimensionality reduction. Despite the evidence of such a connection, few works present theoretical studies regarding redundancy. In this work, we analyze the effect of redundant features on the performance of classification models. We can summarize the contribution of this work as follows: (i) develop a theoretical framework to analyze feature construction and selection, (ii) show that certain properly defined features are redundant but make the data linearly separable, and (iii) propose a formal criterion to validate feature construction methods. The results of experiments suggest that a large number of redundant features can reduce the classification error. The results imply that it is not enough to analyze features solely using criteria that measure the amount of information provided by such features.
YR 2021
FD 2021
LK http://hdl.handle.net/20.500.14066/3777
UL http://hdl.handle.net/20.500.14066/3777
LA eng
NO CONACYT - Consejo Nacional de Ciencia y Tecnología
DS MINDS@UW
RD 03-nov-2024