Modelo matemático combinado para la clasificación de neuroimágenes basado en medidas de similaridad entre hemisferios del cerebro

Abstract

The main contribution of this work is the combination of similarity measures, methods for the construction of subspaces and classification models. Specifically, the NCC was used as a measure of similarity, which was projected to subspace in singular value decomposition following the Eigenfaces methodology, to then apply classification models on these projections. Results with an accuracy of 81% and a predictive capacity of at least 79% were observed for this combination of methods.