Question for Bama Gambler

[ed999]

Hey Bama,

When you used the ASM manual, did you do all the questions in each lesson right after you read the lesson or did you do some of them and came back to the ones you didn't do later on (for reenforcement purposes)? I started on Monday and have just finished lesson 6, Percentile Matching. However I have only did about 40% of the questions following each lesson. I am just curious as to how you used the ASM manual.

Also, do you agree that the Kernel Smoothing lesson was not very well written? Nor do I like the Study Note's approach. I remember taking a class in Survival Analysis back in college last year and that Yellow Book's section on Kernel Smoothing was more understandable.

Thanks!

[Bama Gambler]

Originally Posted by ed999

Hey Bama,

When you used the ASM manual, did you do all the questions in each lesson right after you read the lesson or did you do some of them and came back to the ones you didn't do later on (for reenforcement purposes)? I started on Monday and have just finished lesson 6, Percentile Matching. However I have only did about 40% of the questions following each lesson. I am just curious as to how you used the ASM manual.

Also, do you agree that the Kernel Smoothing lesson was not very well written? Nor do I like the Study Note's approach. I remember taking a class in Survival Analysis back in college last year and that Yellow Book's section on Kernel Smoothing was more understandable.

Thanks!

For some lessons I didn't do all the problems. On some lessons I thought the material was very easy so I didn't feel like I needed to work all the problems. At the very least I would read all the problems. If I felt comfortable that I could work the problem then I would skip it (and come back to it later if I had time).

YES! YES! The Kernel Smoothing lesson and the Study Note's approach on Kernel Smoothing are horrible. I really never got a good grasp of kernel smoothing. I could work the problems but they decided to ask a theory related question on the exam. I missed it.

[Omikron]

This probably won't help Bama much, but if anyone else is looking for an approach to understanding triangular kernel smoothing, I recommend working through the ASM examples while referring back to the formulae in the study notes.

Before you begin please know that for the triangular kernel, observations below the point (in the ASM it was 655) are treated differently than observations above the point. This applies to both the density and the distribution function. Knowing this particular point would have saved me 20 minutes.

Once you see how the formulas work, you probably won't need to memorize them, as Weishaus implies.