4, Part 1. Generate =150 data points as follows: (0,1) , (0,1) , and =0+1+ , where 0=3 and 1=0 . Note that since 1=0 , this means that and are unrelated! Use either the closed form expression for 1^ or a linear regression package of your choice, to obtain the least-squares estimate for 1 . Is it equal to 1 ? Either way, explain. Problem 4, Part 2 Now repeat this experiment 1000 times, always using the same ground truth. You will now have 1000 different values for 1. Plot these values in a histogram (set the bin size so that it looks nice). Note that for each of these 1000 values of 1, you need to generate =150 data points, as you did above. If you do this right, you should have something that looks like a Gaussian. This is telling us that the distribution of the solution to a regression problem looks like a Gaussian. Therefore, we can use everything we know about Gaussians, standard deviations, etc., to judge statistical significance