向量范数
\[||x||_p=(\sum^n_{i=1}|x_i|^p )^{1/p}
\]
距离(Distance、metric)
Minkowski Distance(明可夫斯基距离)
\[||x - y||_p=(\sum^n_{i=1}|x_i - y_i|^p )^{1/p}
\]
manhattan Distance(曼哈顿距离)
当p=1时
\[||x - y||_{p=1}=\sum^n_{i=1}|x_i - y_i|
\]
Euclidean Distance(欧式距离)
当p=2时
\[||x - y||_{p=2}=(\sum^n_{i=1}|x_i - y_i|^2)^{1/2}
\]
Chebyshev Distance(切比雪夫距离)
当\(p=\infin\)时
\[||x - y||_{p=\infin}=max|x_i - y_i|
\]
Mahalanobis Distance(马哈拉诺比斯距离)PCA
\[\sqrt{(x_i-y_i)^T \Sigma^{-1}(x_i-y_i) }
\]
