Each observation has sire as its feature and the output is the price. For example the first house's size is 44m2 and its price is 140TL. ordered the data randomly and then I plcked first 14 houses as my training set, the nert 6 houses (italic ones) as my CV (cross validation) set and the last one for the test set (boid one). Recall that the associated error for any model (both for train, CV and test) is given by the following: J()=2m1i=1m(h0(x(0))y(0))2 Model 1:0=13 and 1=3.5 Model 2:0=3 and 1=3.3 For the training and CV errors, I have this following table For Model1, I have calculated the following value for the first 13 data points in order to calculate the training error: \begin{tabular}{|l|l|} \hline Model 1 & ==113(h(x(0))y(i))2=7069 \\ \hline \end{tabular} For Model2 1 have calculated the following value for the first 5 data points of the CV set (15 to 19) to calculate the CV error: c. (Sp) Which model will you pick? Why? Brleffy pleasel d. (10p) What is the real world performance of the model that you picked in (c)