$($b$)$ pts$)$ True or False: Let $f(x$;$W)$ be a neural network which maps an image, $x,$ to a probability distribution over $10$ classes. Let $W^{***}$ be the parameter matirx$/$tensor learned from a set of training samples. Let $x^{***}$ and $y^{***}$ be a pair of real image and real target. We want to generate a fake image tilde$(x)$ that makes the neural network err, i$.$e$.,$ having wrong prediction. We can compute tilde$(x)$ by:

$($c$)(2$ pts$)$ True or False: The above attack in $($b$)$ is targeted attack.