Accueil

Omer Tamuz

  1. Adaptively Learning the Crowd Kernel.

    Fri, 05/06/2011 - 15:01 — SomNambul
    Authors: Omer Tamuz, Ohad Shamir, Ce Liu, Serge Belongie, Adam Tauman Kalai
    Subjects: Learning
    Abstract

    We introduce an algorithm that, given n objects, learns a similarity matrix
    over all n^2 pairs, from crowdsourced data alone. The algorithm samples
    responses to adaptively chosen triplet-based relative-similarity queries. Each
    query has the form "is object 'a' more similar to 'b' or to 'c'?" and is chosen
    to be maximally informative given the preceding responses. The output is an
    embedding of the objects into Euclidean space (like MDS); we refer to this as
    the "crowd kernel."

  2. Making Consensus Tractable.

    Wed, 07/07/2010 - 15:01 — SomNambul
    Authors: Elchanan Mossel, Omer Tamuz
    Subjects: Statistics
    Abstract

    The process of consensus voting has many distinct advantages: it fosters
    discussion and participation, empowers minorities and independent thinkers, and
    is more likely, after a decision has been made, to secure the participants'
    support for the chosen course of action.

  3. Truthful Fair Division.

    Tue, 03/30/2010 - 15:02 — SomNambul
    Authors: Elchanan Mossel, Omer Tamuz
    Subjects: Computer Science and Game Theory
    Abstract

    We address the problem of fair division, or cake cutting, with the goal of
    finding truthful mechanisms. In the case of a general measure space ("cake")
    and non-atomic, additive individual preference measures - or utilities - we
    show that there exists a truthful "mechanism" which ensures that each of the k
    players gets at least 1/k of the cake. This mechanism also minimizes risk for
    truthful players. Furthermore, in the case where there exist at least two
    different measures we present a different truthful mechanism which ensures that
    each of the players gets more than 1/k of the cake.

  4. Efficient Bayesian Learning in Social Networks with Gaussian Estimators.

    Thu, 02/04/2010 - 16:01 — SomNambul
    Authors: Elchanan Mossel, Omer Tamuz
    Subjects: Applications
    Abstract

    We propose a simple and efficient Bayesian model of iterative learning on
    social networks. This model is efficient in two senses: the process both
    results in an optimal belief, and can be carried out with modest computational
    resources for large networks. This result extends Condorcet's Jury Theorem to
    general social networks, while preserving rationality and computational
    feasibility.

  5. Complete Characterization of Functions Satisfying the Conditions of Arrow's Theorem.

    Fri, 10/16/2009 - 04:10 — SomNambul
    Authors: Elchanan Mossel, Omer Tamuz
    Subjects: Combinatorics
    Abstract

    Arrow's theorem implies that a social choice function satisfying
    Transitivity, the Pareto Principle (Unanimity) and Independence of Irrelevant
    Alternatives (IIA) must be dictatorial. When non-strict preferences are
    allowed, a dictatorial social choice function is defined as a function for
    which there exists a single voter whose strict preferences are followed. This
    definition allows for many different dictatorial functions. In particular, we
    construct examples of dictatorial functions which do not satisfy Transitivity
    and IIA.

Math-arch

Syndicate content