Course talk:CPSC522/Minimax Regret Preference Elicitation for Risky Prospects

From UBC Wiki

Contents

Thread titleRepliesLast modified
Feedback (F3)017:41, 19 March 2019
Peer Review105:17, 19 March 2019

Feedback (F3)

I thought this page was very clear and informative. I just have a couple of very minor comments:

- in Outcome Queries you define s and t as having known utilities and then write them as being in a set called 'unknown'. Should this set be called 'known' instead?

- you have a few places where spaces are missing after math symbols

- the incremental elicitation section could maybe do with a little more explanation. I think all the right information is contained there, but more explanation might make it easier to understand

WilliamHarvey (talk)17:41, 19 March 2019

Peer Review

Hi Ali.

Overall, the page is well written and clearly communicated. I also learned quite a bit about minimax regret from reading it, so thank you.

Some of the math expressions seem incomplete. For example, pairwise max regret notation should have something like PMC(d_i, d_j, C) on left hand side and include brackets for max. Incremental Elicitation section needs some more details as well.

The outcome query segment is not self-explanatory to me. You might want to explicitly define the notations and the variables.

It would be nice to have the plots you mention for the experiments section on the page.

NamHeeKim (talk)16:17, 18 March 2019

Thanks a lot for your comment. I did not explain the sections you mentioned in more detail because they are not part of the incremental contribution of one paper over the other, and therefore, as my understanding of this assignment, not part of the assignment. Those sections exist to make the reader have a more consistent idea of what is actually happening and help keep the page a complete unit as opposed to a union of some pieces that don't fit together. Nevertheless, I will definitely try to expand more on them.

I will definitely fix the math expressions and I will definitely add the plots. Best, Ali

AliMohammadMehr (talk)05:17, 19 March 2019