Why is max($\frac{2}{||w||}$)= min($\frac{1}{2}$)($||w||^2$)? I was watching a video on machine learning.
The instructor says that maximizing ($\frac{2}{||w||}$)is difficult (why?) so instead we prefer to minimize $\frac{1}{2}||w||^2$. $w$ is a vector.
How are these two functions equivalent?
****ADDENDUM ****
I plotted both statements. So I can see that the max of ($\frac{2}{||w||}$) is at 0 (it reaches infinity) and the min of $\frac{1}{2}||w||^2$ is 0. Am I thinking of this correctly? 
Isn't it obvious to begin with? I couldve figured out that the max of ($\frac{2}{||w||}$) is 0 without doing any fancy conversion.

 A: The two expressions are not equivalent.  However,
$$
\frac{2}{\| w \|} \ge \frac{2}{\| x \|}
\quad \Longleftrightarrow \quad
\tfrac{1}{2} \| w \| \le \tfrac{1}{2} \| x \|
\quad \Longleftrightarrow \quad
\tfrac{1}{2} \| w \|^2 \le \tfrac{1}{2} \| x \|^2
$$
All quantities are positive, so we can multiply/divide by them freely.
As to the question of why, the square of the magnitude of a vector is a simpler expression to work with, as it is just the dot product
$$
\| w \|^2 = w \cdot w.
$$
A: $max(\frac{2}{\|w \|})$ is equivalent to $min(\frac{\|w\|}{2})$ with respect to the solution $w$. 
$min(\frac{\|w\|}{2})$ is equivalent to $min(\frac{\|w\|^2}{2})$ with respect to the solution $w$, because they achieve the global optima at the same point. 
A: Let $x = ||w||$, then $\dfrac{2}{||w||} = \dfrac{2}{x}$. So if $\dfrac{2}{x} \leq M$, then:$x \geq \dfrac{2}{M}$, and it follows that: $x^2 \geq \dfrac{4}{M^2}$. This means that instead of finding the max of $\dfrac{2}{x}$ which can be cumbersome, you equivalently solve for the min of $x^2$ which is much simpler because you will deal painlessly with a much simpler object: a polynomial of second degree instead of a painful rational expression !
