Dear all,
You are receiving this email because you have signed up to the mailing
list for announcements regarding the local COMSOC Seminar at the ILLC.
I'm happy to be able to report that we have decided to cautiously
restart the seminar. Our first speaker is Oliviero Nardi, who this
Friday will talk about the work he did for his MSc AI thesis on a
graph-based algorithm for the automated justification of collective
decisions. You will find all details on the seminar website:
https://staff.science.uva.nl/u.endriss/seminar/
I noticed that several people signed up during the pandemic, also people
from outside Amsterdam (and outside the Netherlands). That's of course
fine, but in case you signed up by accident and actually were looking
for the (online, international) COMSOC Video Seminar, let me know
whether you want me to switch you to the other mailing list instead.
All the best,
Ulle
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
I'm delighted to announce the next COMSOC Seminar. This Thursday at
15:00 we'll have Aditya Aradhye from Maastricht University talk about
his work on group-strategyproof aggregation rules in multidimensional
domains. I'm including the abstract below and am looking forward to
seeing you here on Thursday.
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Aditya Aradhye (Maastricht)
Title: Group Strategy-Proof Rules in Multidimensional Domains
Date: Thursday 20 February 2020
Time: 15:00
Location: ILLC Seminar Room F1.15, Science Park 107, Amsterdam
Abstract: We consider a social choice setting in which the alternatives
are binary vectors and the preferences of the agents are determined by
the Hamming distance from their most preferred alternative. We restrict
the set of social choice functions to 'rules' which satisfy unanimity,
anonymity and neutrality. As the preference domain is not the universal
domain, the Gibbard-Satterthwaite Theorem does not hold. Hence, we study
a stronger notion: group strategy-proof rules. We show an impossibility
for strongly group strategy-proof rules.
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
The next COMSOC Seminar will take place this Thursday at 15:30, and the
speaker on this occasion is yours truly. I'm including the abstract at
the end of this message and hope to see you on Thursday.
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Ulle Endriss (ILLC)
Title: Analysis of Matching Mechanisms via SAT Solving
Date and time: Thurday 23 January 2020, 15:30
Location: F1.15, Science Park 107, Amsterdam
Abstract: Matching is a research area at the intersection of game theory
and social choice theory that is concerned with the design of mechanisms
for pairing up agents belonging to two different groups, such as job
seekers and companies. In this talk I will explain how to use ideas from
mathematical logic (model theory) and computational logic (SAT solving)
to obtain impossibility theorems regarding the design of matching
mechanisms that satisfy certain intuitively appealing properties (e.g.,
regarding the stability and fairness of outcomes).
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
Happy new year! The first COMSOC Seminar of the year will be given by
Bernhard von Stengel from the LSE, next week on Thursday at 16:00.
Bernhard will talk about "Game Theory and Politics". This is a first
version of a talk he is developing for a general-interest audience, so
should be fun and easily accessible to everyone (expect lots of Brexit
puns and the like). I'm including the abstract below and hope to see you
next week.
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Bernhard von Stengel (London)
Title: Game Theory and Politics
Date and time: Thursday 16 January 2020, 16:00
Location: F1.15, Science Park 107, Amsterdam
Abstract:
Game theory is the 'science of interaction'. This talk will explain some
insights of game theory and apply them to current politics.
Of course politicians play games. They offer cheap promises that they
think they don't have to fulfil. Such as a "simple" in-out referendum on
EU membership in Britain. That game plan went wrong. Game theory may
have helped, with tools for thinking ahead and concepts of strategy, to
counter the wishful thinking that seems to drive such decisions. Game
theory can also help explain the incentive problems of climate change
and reasons for democratic deadlock. A game-theoretic analysis shows the
importance of the rules of the game, for example in electoral systems.
The aim of the talk is to highlight some uses and mis-uses of game
theory and decision theory with examples from politics, for a general
audience.
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
This Friday at 16:00, Mehmet Ismail from King's College London will
visit us and give a talk at the COMSOC Seminar. You can find an abstract
below. I hope to see you at the seminar.
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Mehmet Ismail (London)
Title: One for all, all for one—von Neumann, Wald, Rawls, and Pareto
Date: Friday 6 December 2019
Time: 16:00
Location: Room F0.25, Science Park 107, Amsterdam
Abstract: Applications of the maximin criterion extend beyond economics
to statistics, computer science, politics, and operations research.
However, the maximin criterion—be it von Neumann's, Wald's, or
Rawls'—draws fierce criticism due to its extremely pessimistic stance. I
propose a novel concept, dubbed the optimin criterion, which is based on
(Pareto) optimizing the worst-case payoffs of tacit agreements. The
optimin criterion generalizes and unifies results in various fields: It
not only coincides with (i) Wald's statistical decision-making criterion
when Nature is antagonistic, but also generalizes (ii) Nash equilibrium
in n-person constant-sum games, (iii) the core in cooperative games,
(iv) stable matchings in matching models, and (v) competitive
equilibrium in the Arrow-Debreu economy. Moreover, every Nash
equilibrium satisfies the optimin criterion in an auxiliary game.
Finally, the optimin criterion is the first parameter-free concept that
can selectively explain the puzzle of cooperation in games, including
the repeated prisoner's dilemma, the traveler's dilemma, the centipede
game, and the repeated public goods game.
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
Next week we are going to have two sessions of the Computational Social
Choice Seminar at the ILLC. On Monday at 15:30 Simon will deliver an
introductory tutorial on participatory budgeting. On Tuesday at 15:30
Zoi will talk about her recent work on the manipulation of positional
scoring rules when voters report incomplete preferences.
I'm including both abstracts below.
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
PS: Apologies for having forgotten to announce Jan Maly's talk to this
mailing list. Jan will still be at the ILLC until the end of next week.
-------------------------------------------------------------------------
Speaker: Simon Rey (ILLC)
Title: A Review of the Computational Social Choice Literature on
Participatory Budgeting
Date and time: Monday 25 November 2019, 15:30
Location: F1.15, Science Park 107, Amsterdam
Abstract: Participatory Budgeting is a growing topic in social choice.
Introduced in the late 1980s in Brazil, it is now viewed as a critical
topic to improve the democratic process. In this talk I will introduce
the concept of Participatory Budgeting and give an overview of the
related literature in the field of Computational Social Choice.
-------------------------------------------------------------------------
Speaker: Zoi Terzopoulou (ILLC)
Title: Strategic Manipulation with Incomplete Preferences
Date and time: Tuesday 26 November 2019, 15:30
Location: F3.20, Science Park 107, Amsterdam
Abstract: Many websites that recommend various services use
crowdsourcing to collect reviews and rankings. These rankings, usually
concerning a subset of all the offered alternatives, are then
aggregated. Motivated by and generalising upon such scenarios, we
axiomatise a family of positional scoring rules for profiles of possibly
incomplete individual preferences. But many opportunities arise for the
agents to manipulate the outcome in this setting. They may lie in order
to obtain a better result by: (i) switching the order of a ranked pair
of alternatives, (ii) omitting some of their truthful preferences, or
(iii) reporting more preferences than the ones they truthfully hold.
After formalising these new concepts, we characterise all positional
scoring rules that are immune to manipulation. This talk is based on is
joint work with Justin Kruger.
-------------------------------------------------------------------------
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
The next talk in the Computational Social Choice Seminar at the ILLC
will be given on Tuesday by Simon Rey, who started his PhD at the ILLC a
few weeks ago. Simon will be speaking about fair division in the
presence of both goods and bads. Hope to see you there!
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Simon Rey (ILLC)
Title: Almost Group Envy-free Allocation of Indivisible Goods and Chores
Date and time: Tuesday 5 November 2019, 16:00
Location: F3.20, Science Park 107, Amsterdam
Abstract: In this talk I will present some recent work I have been doing
with Haris Aziz. We consider a multiagent resource allocation setting in
which an agent's utility may decrease or increase when an item is
allocated. We present stronger and relaxed versions of the group
envy-freeness concept that are especially suitable for the allocation of
indivisible items. Of particular interest is a concept called group
envy-freeness up to one item (GEF1). We study which fairness concepts
guarantee the existence of a fair allocation under which preference
domain. For two natural classes of additive utilities, we design
polynomial-time algorithms to compute a GEF1 allocation.
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
I hope you had an enjoyable summer. The Computational Social Choice
Seminar at the ILLC is back in session now. We'll start with a talk by
Aidan Lyon this Friday at 16:00. Aidan is CEO and co-founder of
DelphiCloud, a start-up here at the Science Park. See abstract below. I
hope to see you this Friday!
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Aidan Lyon (DelphiCloud)
Title: DelphiCloud: Eliciting and Aggregating Expert Estimates
Date: Friday 13 September 2019
Time: 16:00
Location: Room F1.15, Science Park 107, Amsterdam
Abstract: In this talk I give a brief overview of the business and
research work of DelphiCloud, a company created by philosophers for the
purpose of bringing philosophical and psychological research to the
private sector. I'll then present in detail some of the work we are
doing with the repliCATS team at the University of Melbourne as part of
the DARPA SCORE project. The SCORE project is an attempt to address the
reproducibility crisis in the social sciences in part by getting experts
to estimate the reproducibility of thousands of studies in these fields.
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
The next Computational Social Choice Seminar at the ILLC will take place
on Monday at 15:00. It will be delivered by Lefteris Kirousis (Athens),
who is going to be visiting for the day. He will speak about possibility
results in judgment aggregation. I'm including the abstract below.
As always, for more information on the COMSOC Seminar please consult
http://www.illc.uva.nl/~ulle/seminar/.
All the best,
Ulle
-------------------------------------------------------------------------
Speaker: Lefteris Kirousis (Athens)
Title: Abstract Possibility Domains: Algorithms and Characterizations
Date: Monday 1 July 2019
Time: 15:00
Location: ILLC Room F1.15, Science Park 107, Amsterdam
Abstract
Let V be a set of possible evaluations on an issue, with cardinality two
or possibly greater. An abstract domain (or just domain) is an arbitrary
subset D of a Cartesian power V^m. The elements of a domain are
considered to represent the rational, or permissible, evaluation vectors
on m issues, in some abstract sense of rationality. Given an integer
k>1, a k-ary aggregator for D is a function F defined for all elements
of D^k and taking values in D. We assume that aggregators are defined
independently on each issue and that they are supportive. An aggregator
F is called non-dictatorial if it differs from all projection functions
on any d in 1..k. A domain D is called a possibility domain if it admits
a non-dictatorial aggregator of some arity k. I will first give
necessary and sufficient conditions for D to be a possibility domain in
terms of existence of binary or ternary aggregators. Using this
characterization as a stepping stone, I will give efficient (polynomial
time in the size of the domain) algorithms that decide whether an input
domain is a possibility one. Finally in the case the cardinality of V is
exactly two (Boolean case), I will define a class of Conjunctive Normal
Form formulas of Propositional Logic that describe the domains that are
possibility ones. I will also prove that such formulas are recognizable
in linear time in the size of the input formula.
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/
Dear all,
Unfortunately, we had to cancel the talk by Frank Feys in the COMSOC
Seminar at the ILLC planned for this Tuesday, because Frank got sick. I
hope that we will be able to reschedule his talk sometime soon.
http://www.illc.uva.nl/~ulle/seminar/
All the best,
Ulle
--
Ulle Endriss
ILLC, University of Amsterdam
http://www.illc.uva.nl/~ulle/