This page looks at all the problems I missed on the midterm, analyzes why I got them wrong, and what I can do in the future to make sure I do better.
I got this problem wrong because of a rounding issue. I cut off the number of decimal points I returned for the test results and the answer was rounded up to 0.602. I attended office hours and after talking with Professor Frazier I think the way to solve this is to not automate the rounding, but to print the whole value and then leave the rounding up to me.
Similar problem here as with 18, I got this problem wrong because of a rounding issue. I cut off the number of decimal points I returned for the test results and the answer was rounded up to 0.602. I attended office hours and after talking with Professor Frazier I think the way to solve this is to not automate the rounding, but to print the whole value and then leave the rounding up to me.
Same problem as 18 and 19, I got this problem wrong because of a rounding issue. I cut off the number of decimal points I returned for the test results and the answer was rounded up to 0.602. I attended office hours and after talking with Professor Frazier I think the way to solve this is to not automate the rounding, but to print the whole value and then leave the rounding up to me.
I got this question wrong partly because I didn’t understand what data it was asking for, but also because I didn’t understand what to return. I thought it was asking for the least correlated feature from problem 15 and 16, and I thought I was supposed to return the r2 value of the model that had the lowest score.
Instead, I was supposed to use the whole standardized dataset, create the three models with the previously found hyperparameters, fit the created models, and return the coefficient for the least correlated feature.
I got this question wrong almost exactly the way I got question 21 wrong. I didn’t understand what data it was asking for or what to return. I thought it was asking for the most correlated feature from problem 15 and 16, and I thought I was supposed to return the r2 value of the model that had the lowest score.
Instead, I was supposed to use the whole standardized dataset, create the three models with the previously found hyperparameters, fit the created models, and return the coefficient for the most correlated feature.
The issue for this answer is near identical to the issue for 18, 19, and 20. I entered the answer as 0.002, because of the automated rounding: print('Optimal alpha value: ' + format(a_range[idx], '.3f')). If I had left off the rounding, I would have gotten 0.00186, which is the correct answer.