The functions $f:\mathbb{R}^n\longrightarrow \mathbb{R}^k$ and $g:\mathbb{R}^k\longrightarrow \mathbb{R}^m$ can be composed to the function $g\circ f: \mathbb{R}^n\longrightarrow \mathbb{R}^m$. We have the following chain rule for differentiation.