Spatial and Anatomical Regularization Based on Multiple Kernel Learning for Neuroimaging Classification

YingJiang WU  BenYong LIU  

IEICE TRANSACTIONS on Information and Systems   Vol.E99-D   No.4   pp.1272-1274
Publication Date: 2016/04/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2015EDL8163
Type of Manuscript: LETTER
Category: Biological Engineering
neuroimaging,  spatial regularization,  anatomical regularization,  multiple kernel learning,  

Full Text: PDF>>
Buy this Article

Recently, a high dimensional classification framework has been proposed to introduce spatial and anatomical priors in classical single kernel support vector machine optimization scheme, wherein the sequential minimal optimization (SMO) training algorithm is adopted, for brain image analysis. However, to satisfy the optimization conditions required in the single kernel case, it is unreasonably assumed that the spatial regularization parameter is equal to the anatomical one. In this letter, this approach is improved by combining SMO algorithm with multiple kernel learning to avoid that assumption and optimally estimate two parameters. The improvement is comparably demonstrated by experimental results on classification of Alzheimer patients and elderly controls.