Who can see your viewing activity?
QUESTION: Like this?
QUESTION: what is "dom" parameter?
QUESTION: what if we bound some more restrictive parameter - say vertex cover -- can we then have undouded domain size and fet FPT?
QUESTION: can you tell something about the online model where agents arrive over time?
Ok thank you!
QUESTION: Is there any benefit of *not* using the Shapley solution for individual taxis in order to get fairness/stability?
@Haris Thanks for your question! This may be a silly answer but if we require only one agent to pay the total cost, some counter examples for EF existence does not work.
:) ok thanks
@Ayumi: what about applying Shapley value on the game induced by multiple taxis?
QUESTION: are the percentage values from exhaustive enumeration of profiles or does it come from simulations?
Q: Can you explain a bit about the linear program you used?
QUESTION: Why are there more manipulations possible with a single manipulator than with a coalition?
I must have misread the table.
I Don't think so
Computing volumes is really lonked to the IAC assumption
QUESTION: Do you think this could be used as a source of realistic data for approval elections (by ignoring the costs)?
just a penny for your though -- I will not hold you responsible :)
QUESTION: Can you give an example of research question that would be easy to address using your data library? What makes it different from other repositories such as Prefilb?
@Paolo: Evaluate the running time of an ILP solver for a PB rule on realistic data
@Paolo About the difference with Preflib, in Preflib we do not have data with cost for instance. It is just a different type of data.
Also, Staszek is in a unit that has long tradition of getting full voting data for Polish elections, so they have quite some credibility
Paul, does it make sense to look at single-peaked electorates in this setting?
QUESTION: Can this be phrased as (a restricted version of) a strategic candidacy game?
Could you estimate the number of profiles where a NE exists ? Is it rare ?
QUESTION: Did you/are you planning to consider a 2D case (e.g. a political compass kind of scenario) with some metric (e.g. Euclidean)?
Markus, yes. Having finitely many alternatives does change NE existence
QUESTION: Is the case when all the candidates of a party form an interval a reasonable consideration as a tractable subclass? (The instance based on 3-SAT did not appear to be an instance of this kind.)
This is for the case of more than two parties...
QUESITON : What if we measure the voter satisfaction in average ? Is there a difference in the gurentees between two parties to multi parties ?
Good talk Paul :)
I meant that the “hard instances” did not seem to be of the interval kind - but I see that interval case + 2 parties is a nice situation, thanks! :)
Very interesting potential direction, thanks
@ Markus and Paul: I see it as a generalisation of strategic candidacy. SC is obtained by (1) having one party per potential candidate c, with two ‘actual candidates’ who are c and an artificial candidate who is going to gain zero vote and thus have no effect on the election. Does it make sense?
@Jérôme: Thank you for the suggestion. That seems to make perfect sense! Interesting!
Does a DiRe committee always exist?
QUESTION: there is a lot of writing on fairness both in behavioral psychology (Kahnemann) and in moral psychology (Haidt). Can you relate these notions to discussions from these fields?
What about DiRe committee and the existence of a Condorcet committee?
Can you make the European countries on the diifferent rules, using irs voting mecanims ?
Q: Is your work actually used by the EU?
There are other issues where the EU had a division problem, e.g., migrants. Can you framework be applied to that?