Table 4 Some choices for obtaining w
From: Bi-linear matrix-variate analyses, integrative hypothesis tests, and case-control studies
Choice | Description |
|---|---|
(a) | get w via FDA by Equation (11), as addressed in the previous subsection. |
(b) | estimate w by maximizing L by Equation (4), as to be addressed in the next subsection. |
(c) | get w as the normal direction of a separating hyperplane by one of machine learning approaches, e.g., support vector machine (SVM) (Cortes and Vapnik 1995; Suykens et al. 2002). |