Midterm exams Q10

I used excel to compute the line. I cheated :(. It helps to turn on the caption in the videos.

I don't remember what i was doing at your age. Definitely not artificial intelligence. Maybe mario brothers or contra.

Linear regression was covered in exactly the same form in the homework. So assuming you understood the homework fully, you should be able to answer this question easily.

Having said that... linear regression is indeed quite hard for someone born in 2000. On the other hand, it's too difficult for the vast majority of folks in their 20's, 30's, 40's, or 50's too.

I don't see how the professors could modify this question to be more accessible to younger students. It's difficult, but it follows directly from the material presented in class and homework.

With linear regression, you have to compute various "sums", and then simply plug them into the final formula. You can study this elsewhere (just Google linear regression).

One habit that's worth getting into from an early age is that of double-checking your answers by an independent means, wherever possible.

Here you can easily do that by taking your calculated coefficients and then plugging in the values you're given for the Xi (1,3,4,5,9) and seeing what values it calculates for the corresponding Yi. If you've made a mistake, or the algorithm has not performed well (Prof. Thrun touches on this in the lecture video), the calculated Yi values will be different to those that were given and which you used to compute the linear regression coefficients.

Another way to achieve this double-check is to graph the data you're given, and then on the same graph plot the straight line you calculated using linear regression. The line should follow the data, more or less: if not, there's trouble that needs to be sorted out!

In the exam question the linear regression analysis is perfect, meaning that the data follows the line calculated through standard linear regression precisely. This is rarely the case, but here it gives you a unique opportunity to be absolutely certain about your answers before even entering them.

Personally, I don't think age has anything to do with the ability to learn. All you need to know for this class is the prerequisite knowledge. This class required background knowledge, and if you know how to perform a sum, then you can solve question 10, a linear regression (least squares). You may not understand the derivation because you aren't familiar with calculus and summations, but they were prerequisites for this class.

You don't need to remember the formulae in this class, because you can always access them. Try collecting them into a 'cheat sheet' for a quick reference. There are many mnemonics for remembering the formula for regression, especially if you know matrices.

I speak of course of this image

Weisstein, Eric W. "Least Squares Fitting." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/LeastSquaresFitting.html

i love Ai but somethings are hard for me but Im willing to learn .atleast IM trying .