# Looking for Good Summary Lectures in Simple&Multiple Linear Regression with Matrices

May I say that my instructor for my upper-year regression analysis course in ineffective and insufficient to say the least...due to privacy concerns I will not provide screenshot of the slides.

I am looking for in-depth, clearly explained material with examples, especially in SLR and MLR working with matrices (projection matrices, etc.). The level is upper-year Undergraduate / graduate. I looked at STA501 from Penn State. Doesn't have enough info, as seen from their model assumptions. My suggested text is "Regression Analysis by Example, 5th Edition". 1).If there are solutions to the problems it would be great.

2).Condensed "cheat sheet" / information sheet that's not Intro stats and not Computer Science intro linear regression is highly sought after. I do have one I made, but it is disorganized, not very accurate, and is probably not going to really help. While I know that the midterm is not a memory test (that's why a cheat sheet is allowed), how do I ensure that I remember to put all info that I didn't even know to put on the sheet and/or stuff that I would've forgotten during the test, on the information sheet?? And yes, the midterm is going to be tricky, and will contain proofs :( For example, we need to know these once-proofs, now facts, to solve other problems, like P' = P, P^2 = P, (I-P)X = 0, etc.

3).If there are good practice problems with answers it might just help me out. (I've already looked at http://users.stat.ufl.edu/~mripol/3024/PracticeExamRegression3024.pdf but it lacked working with matrices)

4). Some example questions are like this one: Multiple Linear Regression and Correlation of two beta estimates but I can only guess the answer since (2,2) and (4,4) are the same. What if it asks for the estimated correlation between beta 2 and beta 3? What does the rows and columns represent in (X'X)^-1

Thanks.