Survival Data Analysis
BIOS 7210 / STAT 7570
Fall 2017
Note: There is a newer version of this course (Fall 2019) available here.
Announcements:
9-13: For several reasons, Exam I is going to be pushed back a week. Our first exam will now take place on October 5; the schedule has been updated to reflect this.
9-08: I made a few revisions/edits to the slides for the 9-07 lecture to clarify the notation, which turned out to be a little confusing during our derivations in class.
8-14: Welcome to the website for Survival Data Analysis. On the menu above are links to the syllabus, lecture notes, and assignments. There are also some data sets available in a tab-delimited format (we will be adding data sets to this page over the course of the semester). We will discuss these data in class and on homework assignments as the semester goes on.
Corrections:
11-10: This is way overdue, I realize, but I revised slides 17-19 of the 10-12 notes where we derived the likelihood and score function for the log-linear AFT model. As Monica pointed out at the time, the notation I chose was a bit problematic, as in some sense we have two likelihoods, one with respect to Y and the other with respect to W, that differ by a \( \sigma \) term that comes from the Jacobian. My original notation was confusing and vague on this point; the new slides are much better. Also, as Jun pointed out to me, this has several implications for Problem 1 on Assignment 8, which I also revised according to the new notation. Sorry about the confusion – thank you very much to Monica and Jun for pointing out these issues.
10-25: As several students pointed out, the notation I chose for the 10-24 lecture wasn’t ideal, in that it used the indices i and j in a different way than we had been using them in previous lectures. The lecture has now been rewritten so that the indexing notation is more consistent. I think this makes the lecture easier to follow, but if you prefer the original, here it is.
9-28 – Lecture 9-21, Slide 15: The proposition should read \( \ell(\theta) \approx \ell(\hat{\theta}) - \frac{1}{2}I(\hat{\theta})(\theta - \hat{\theta})^2 \). We derived this correctly on the board in class, but the slides had a \(+\) instead of a \(-\). Thank you to Yiheng for pointing this out.
8-26 – Lecture 8-24, Slide 21: In the last line of the equation, \( t_j < t \) should be \( t_j \leq t \). Thank you to Yiheng for pointing this out.