Disclaimer: Dieser Thread wurde aus dem alten Forum importiert. Daher werden eventuell nicht alle Formatierungen richtig angezeigt. Der ursprüngliche Thread beginnt im zweiten Post dieses Threads.
Q&A notes thread
Dear students,
I will be posting notes from the Q&A lecture sessions in this thread. The first two Q&A were pretty quiet. There were two highlights that are worth mentioning:
-
If there is a large number of participants in a call, it is much easier to use the feedback from the reactions in the tab “Participants” than from “Chat” (especially for raising hands, yes and no responses, etc). You can find the buttons to open these two tabs just under the main video stream. The side-by-side mode (in the “View Options” menu, above the main video) can also be helpful.
-
You could think of the lecture videos and Q&A session as a case of a flipped classroom. It will help you retain focus if you try to come up with a question while you are watching the lectures - and then of course, ask it in the Q&A. By the way, this is also a good academic practice when you go to conferences
Unfortunately, I’m not quite able to keep up with the notes from the Q&A sessions. I would recommend attending them, even if you just come to lurk. Asking questions is not mandatory
The question of notation comes up every now and then, so here is a quick summary of the conventions we try to keep to.
In this course, we are only considering random variables with finite domains, often boolean random variables.
Example: P(Weather=sunny) (prior probability, or unconditional probability). The variable name is written with a capital letter.
-
Boolean random variables: [m]A[/m], [m]B[/m]. For a Boolean variable Name, we write [m]name[/m] for Name = T and [m]¬name[/m] for Name = F.
-
General finite-domain random variables: [m]X[/m], [m]Y[/m]
-
The probability distribution for a random variable [m]X[/m], written [b]P/b (bold P), is the vector of probabilities for the (ordered) domain of [m]X[/m].
Example: P(Headache) = <0.1, 0.9> -
Set of random variables: Z (name of set in bold).
-
Joint probability [b]P/b - probabilities of all events for all variables in the set Z.
This is from a while back. In the lecture notes, the preference relations is written as A ≻ B, meaning A is preferred over B. In the lecture videos, this was the other way around. The version in the current notes is correct.