Through maximalizing the margin, we can obtain the optimal projection vector, and avoid the small sample size problem due to singularity of the within-class scatter.
通过极大化该边界获得最优投影向量,同时避免因类内离散度矩阵奇异导致的小样本问题。
Through maximalizing the margin, we can obtain the optimal projection vector, and avoid the small sample size problem due to singularity of the within-class scatter.
通过极大化该边界获得最优投影向量,同时避免因类内离散度矩阵奇异导致的小样本问题。
应用推荐