Presenting Theorems for Exchangeable Binary Random Variables with Applications to Expert Groups and Voting Power at 12:15-13:45 in QA406.
On 28th November 2018 we have Serguei Kaniovski of the Austrian Institute of Economic Research (WIFO) visiting us. He is going to give a seminar titled Theorems for Exchangeable Binary Random Variables with Applications to Expert Groups and Voting Power at 12:15-13:45 in room A406 in Building Q, Budapest University of Technology and Economics, Faculty of Economic and Social Sciences, Magyar tudósok körútja 2, 1117 Budapest.
Abstract: The presentation discusses (a) parameterizations of the joint probability distribution of correlated binary random variables, (b) the probability of at least k successes in n exchangeable correlated binary trials and (c) its bounds when the correlations are unknown. This probability finds wide application in reliability and decision theory. It can be used to compute the probability of an expert group collectively reaching the correct decision (Condorcet Jury Theorem) and voting power of each expert as the probability of her casting a decisive vote (Penrose-Banzhaf Model) when the votes are correlated. Empirical evidence refutes the assumption of independent votes required in the classic versions of the Condorcet Jury Theorem and the Penrose – Banzhaf measure of voting power.