Cauchy Schwarz Inequality for Vectors - Quadratic form Clarification

105 Views Asked by At

In one of the proofs of Cauchys Inequality that reduces vector inner product to a quadratic form:

$$0 \leq \langle u + \alpha, u + \alpha \rangle = \langle u , u \rangle + 2 \alpha\langle u ,v \rangle + \alpha^2\langle v,v \rangle$$

Since any quadratic $$ a \alpha^2+2b\alpha + c $$ takes its minimal value @ $ c - \frac{b^2}{a} $ when $ \alpha = -\frac{b}{2a}$ ---- (1)

and the inequality should hold for even this minimum value of the polynomial $$ 0 \leq \langle u,u \rangle - \frac {\langle u,v\rangle^2}{\langle v,v \rangle} \iff \frac {|\langle u,v \rangle|}{||u|| ||v||} \leq 1 $$ --- (2)

Questions:

Main Q) Am not confident how the equivalence is same. Please confirm the following the is correct thinking: $$ 0 \leq \langle u,u \rangle - \frac {\langle u,v\rangle^2}{\langle v,v \rangle} $$ Take the - to the left, we get: $$ \frac {\langle u,v\rangle^2}{\langle v,v \rangle} \leq \langle u,u \rangle $$ Since in vector inner product of the same vector $\langle v,v \rangle$ = $|v|^2$ , thus: $$ \frac {\langle u,v\rangle^2}{|v|^2} \leq |u|^2 $$ $$ \frac {\langle u,v\rangle^2}{|v|^2|u|^2} \leq 1 $$ Taking Square root on both sides: $$ \frac {\langle u,v\rangle}{|v||u|} \leq 1 $$

Would the above be a correct mathematical deduction of how the equivalence is achieved?

Side Q) in (1) Is the minima of a Quadratic equation when its discriminant = 0? Any link that helps show this please suggest.