A structural similarity kernel is presented in this paper for SVM learning, especially for learning with imbalanced datasets. Kernels in SVM are usually pairwise, comparing the sim...
Linear and Quadratic Discriminant Analysis have been used widely in many areas of data mining, machine learning, and bioinformatics. Friedman proposed a compromise between Linear ...
—While medical imaging typically provides massive amounts of data, the extraction of relevant information for predictive diagnosis remains a difficult challenge. Functional MRI ...
This paper introduces a novel regularization strategy to address the generalization issues for large-margin classifiers from the Empirical Risk Minimization (ERM) perspective. Fi...
Dwi Sianto Mansjur, Ted S. Wada, Biing-Hwang Juang
Active shape model (ASM) statistically represents a shape by a set of well-defined landmark points and models object variations using principal component analysis (PCA). However, ...