对称矩阵最小特征值表达式

设$\mathbf{A}\in {\mathbf{S}}^n$
$${\lambda }_{min}\left(\mathbf{A}\right)=\underset{\genfrac{}{}{0ex}{}{\mathbf{Y}⪰0}{\mathrm{tr}\left(\mathbf{Y}\right)=1}}{\min }⟨\mathbf{Y},\mathbf{A}⟩$$
其中${\lambda }_{min}$代表最小特征值
$⟨\mathbf{Y},\mathbf{A}⟩=\mathrm{tr}\left({\mathbf{Y}}^T\mathbf{A}\right)$

证明：
因为$\mathbf{Y},\mathbf{A}\in {\mathbf{S}}^n$
对其进行谱分解
$$\mathbf{A}=\sum _{i=1}^n{\lambda }_i{\mathbf{p}}_i{\mathbf{p}}_i^T$$
$$\mathbf{Y}=\sum _{i=1}^n{\mu }_i{\mathbf{q}}_i{\mathbf{q}}_i^T$$
其中${\lambda }_i$为$\mathbf{A}$的特征值，${\mathbf{p}}_i$为${\lambda }_i$对应的特征向量，${\mathbf{p}}_i$两两标准正交
${\mu }_i$为$\mathbf{Y}$的特征值，${\mathbf{q}}_i$为${\mu }_i$对应的特征向量，${\mathbf{q}}_i$两两标准正交
$$\mathrm{tr}\left(\mathbf{Y}\right)=1\Rightarrow \sum _{i=1}^n{\mu }_i=1$$
$$\begin{array}{rl}& ⟨\mathbf{Y},\mathbf{A}⟩\\ =& \mathrm{tr}\left({\mathbf{Y}}^T\mathbf{A}\right)\\ =& \mathrm{tr}\left(\mathbf{YA}\right)\\ =& \mathrm{tr}\left(\sum _{i=1}^n{\mu }_i{\mathbf{q}}_i{\mathbf{q}}_i^T\mathbf{A}\right)\\ =& \mathrm{tr}\left(\sum _{i=1}^n{\mu }_i{\mathbf{q}}_i^T\mathbf{A}{\mathbf{q}}_i\right)\\ =& \sum _{i=1}^n{\mu }_i{\mathbf{q}}_i^T\mathbf{A}{\mathbf{q}}_i\\ =& \sum _{i=1}^n{\mu }_i{\mathbf{q}}_i^T\left(\sum _{j=1}^n{\lambda }_j{\mathbf{p}}_j{\mathbf{p}}_j^T\right){\mathbf{q}}_i\\ =& \sum _{i=1}^n\sum _{j=1}^n{\mu }_i{\mathbf{q}}_i^T{\lambda }_j{\mathbf{p}}_j{\mathbf{p}}_j^T{\mathbf{q}}_i\\ =& \sum _{i=1}^n\sum _{j=1}^n{\mu }_i{\lambda }_j{\left({\mathbf{q}}_i^T{\mathbf{p}}_j\right)}^2\\ =& \sum _{j=1}^n{\lambda }_j\sum _{i=1}^n{\mu }_i{\left({\mathbf{q}}_i^T{\mathbf{p}}_j\right)}^2\\ \ge & {\lambda }_{min}\sum _{j=1}^n\sum _{i=1}^n{\mu }_i{\left({\mathbf{q}}_i^T{\mathbf{p}}_j\right)}^2\\ =& {\lambda }_{min}\sum _{i=1}^n{\mu }_i\sum _{j=1}^n{\left({\mathbf{q}}_i^T{\mathbf{p}}_j\right)}^2\\ =& {\lambda }_{min}\sum _{i=1}^n{\mu }_i{\mathbf{q}}_i^T\mathbf{P}{\mathbf{P}}^T{\mathbf{q}}_i\\ =& {\lambda }_{min}\sum _{i=1}^n{\mu }_i{\mathbf{q}}_i^T{\mathbf{q}}_i\\ =& {\lambda }_{min}\sum _{i=1}^n{\mu }_i\\ =& {\lambda }_{min}\end{array}$$