PCA
Unsupervised, maximize the variance
LDA
Supervised， minimize distance between same class, maximize different
with-in class scatter, for each class $S_i$ and correspond mean vector $m_i$
$$S_i=\sum (x-m_i)(x-m_i)^T$$
Lycheeee
0 声望1 粉丝
PCA
Unsupervised, maximize the variance
LDA
Supervised， minimize distance between same class, maximize different
with-in class scatter, for each class $S_i$ and correspond mean vector $m_i$
$$S_i=\sum (x-m_i)(x-m_i)^T$$
**粗体** _斜体_ [链接](http://example.com) `代码` - 列表 > 引用。你还可以使用@来通知其他用户。