1. [30 marks] (Linear Regression) Linear regression aims to learn the parameters from the training set D = {(700),y(i)), i = 1,2,..., m} so that the hypothesis hom) = T . i can predict the output y given a new input vector i. Please derive the stochastic gradient descent update rule, that is to use stochastic gradient descent algorithm to update so as to minimize the least squares cost function J (7)