What is the loss function for linear regression-Gradient Descent? 

How does gradient descent update the estimate, give the general formulae?

Given X = [1, 2, 3, 4], and Y = [1.5, 2, 2.5, 5). If we start with 0,=0.5 and 0,=0.5, compute , and after the first iteration if the learning rate is 0.23. Show all your steps, work and calculation.