张量积难点探讨：基于泛性质的多线性映射转化推导

Alright, let's walk through this derivation assuming we already get the core intuition behind tensor products—they're our go-to tool for translating multilinear algebra problems into linear algebra ones—and we're familiar with the relevant mapping notation.

Starting from the universal property of tensor products, we have this foundational equivalence:
$$B(v,w) = L(v\otimes w)$$

We can unpack this step-by-step to get to the final boxed equality:

• First, substitute any vectors $v$ and $w$ with their expansions in a basis: $v = \sum v^i\vec{e}_i$ and $w = \sum w^j\vec{e}_j$. Plugging these into both sides gives us:
  $$B(\sum v^i\vec{e}_i,\sum w^j\vec{e}_j) = L(\sum v^i\vec{e}_i\otimes \sum w^j\vec{e}_j )$$

• Next, leverage the multilinearity of $B$ (it's linear in each input) and the linearity of $L$ (it preserves linear combinations in the tensor product space) to pull the sums outside the mappings. This expands both sides to:
  $$\boxed{\sum_{i}\sum_{j} v^iwjB(\vec{e}i,\vec{e}j) =\sum{i}\sum{j} v^iwj L(\vec{e}_i\otimes \vec{e}_j)}$$

The big takeaway here is how tensor products let us reduce a multilinear problem (dealing with $B$) to a linear one (dealing with $L$) by breaking everything down to the basis level—exactly why we love using tensor products as a translation layer between these two algebraic frameworks.

内容的提问来源于stack exchange，提问作者M.N.Raia