M. Santos Pascual, D. Ríos, P. Ángulo
The Bradley–Terry model is a widely used framework for describing the probabilities of possible outcomes that need to be compared. The model presented in this document takes advantage of this framework, providing a basis for determining which products to display based on user preferences. By combining a Plackett–Luce model with a multinomial logit model, together with a Bayesian inference scheme involving Sequential Hamiltonian Monte Carlo, this study aims to make inferences and predictions about consumers' preferences. It is illustrated in the domain of multimodal transportation route choice and describes how it is combined with a multicriteria route generation scheme.
Keywords: Hamiltonian Monte Carlos, Placket-Lucce model, Bayesian Inference
Scheduled
Young Researchers in Bayesian Statistics
June 10, 2025 7:00 PM
Sala de prensa (MR 13)