3. Regression using Huber Loss. In the class, we defined the Huber loss as lo (e) = if lel d Consider the robust regression model N min ['( yi - 0 xi), i=1 where a; and y; denote the i-th input sample and output/response, respectively and O is the un- known parameter vector. i) Provide the steps of the batch gradient descent in order to obtain the solution for 0. ii) Provide the steps of the stochastic gradient descent using mini-batches of size 1, i.e., one sample in each mini-batch, in order to obtain the solution for 0