深度学习

Conv2d 实现

源码： 源码

Conv = unfold + maymul + fold

unfold函数理解

jupyter 文件

Curse of dimensionality

As the number of feature grows, the amount of data we need to generalise accurately grows exponentially.

1. 覆盖特征值范围20%所需的训练样本数量随着维度增加呈指数型增长.
2. 维度d很大时，超球面的体积趋于零.