求翻译：PCA的问题其实是一个基的变换，使得变换后的数据有着最大的方差。方差的大小描述的是一个变量的信息量，我们在讲一个东西的稳定性的时候，往往说要减小方差，如果一个模型的方差很大，那就说明模型不稳定了。但是对于我们用于机器学习的数据（主要是训练数据），方差大才有意义，不然输入的数据都是同一个点，那方差就为0了，这样输入的多个数据就等同于一个数据了。是什么意思？

null

The issue of PCA is actually a transformation of the base, making the data after the transformation have the greatest variance. The size of the variance describes a variable amount of information, we are talking about the stuff of stability, they often said that to reduce variance, if a model the va

The PCA question is actually a base transformation, after enables the transformation the data to have the biggest variance.The variance size description is a variable information content, we are speaking a thing stable time, often said must reduce the variance, if a model variance is very big, that

Problem is actually a radical transformation of the PCA, which makes the data with the largest variance. Is described by the size of a variable amount of variance, when we tell a stability, often said to reduce the variance, if one models the variance is large, it indicates that the model is not sta

null