New Flexible Probability distributions for ranking data

Risultato della ricerca: Other

Abstract

Recently, several models have been proposed in literature for analyzingranks assigned by people to some object. These models summarize the liking feelingfor this object, possibly also with respect to a set of explanatory variables. Somerecent works have suggested the use of the Shifted Binomial and of the Inverse Hypergeometricdistribution for modelling the approval rate, while mixture modelshave been developed for taking into account the uncertainty of the ranking process.We propose two new probabilistic models, based on the Discrete Beta and theShifted-Beta Binomial distributions, that ensure much flexibility and allow the jointmodelling of the scale (approval rate) and the shape (uncertainty) of the distributionof the ranks assigned to the object.
Lingua originaleEnglish
Stato di pubblicazionePublished - 2013

    Fingerprint

Cita questo