Numerical analysis and image processing (from a DSP perspective) went okay. Stats and advanced linear algebra less so.
I forgot the bounds that prove that an MLE is a minimum variance unbiased estimator when it is unbiased, in the stats case[1], and I didn't know the eigenvectors and -values for a particular type of matrix in the latter, nor how to augment certain root-finding methods to respect equality constraints at every step.
[1] I understood that was what I was supposed to have written there when there were 15 seconds or so left and I was separating the carbon copy pages. Similarly to how it's much more annoying to miss a train by half a minute than by half an hour, it's much more annoying to do something wrong and realize it just as it's too late than realize it days later. But I suppose the point is that if I knew the stuff better, I would have recalled it in time to begin with.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
branchandroot
no subject
Date: 2015-06-14 09:43 pm (UTC)I forgot the bounds that prove that an MLE is a minimum variance unbiased estimator when it is unbiased, in the stats case[1], and I didn't know the eigenvectors and -values for a particular type of matrix in the latter, nor how to augment certain root-finding methods to respect equality constraints at every step.
[1] I understood that was what I was supposed to have written there when there were 15 seconds or so left and I was separating the carbon copy pages. Similarly to how it's much more annoying to miss a train by half a minute than by half an hour, it's much more annoying to do something wrong and realize it just as it's too late than realize it days later. But I suppose the point is that if I knew the stuff better, I would have recalled it in time to begin with.