PLS, Linear Models and Invariant Spaces
从高斯-马尔可夫模型视角审视偏最小二乘法，利用不变子空间阐明其原理，并给出与主成分回归的联系。
Partial least squares (PLS) is considered from the perspective of the Gauss- Markov model. Linear spaces play a fundamental role and, in particular, invariant subspaces serve to illuminate the PLS method. Some connections between PCR and PLS are also given.