Set values related to the prior distributions for the Bayesian Mallows model.
Usage
set_priors(gamma = 1, lambda = 0.001, psi = 10, kappa = c(1, 3))Arguments
- gamma
Strictly positive numeric value specifying the shape parameter of the gamma prior distribution of \(\alpha\). Defaults to
1, thus recovering the exponential prior distribution used by (Vitelli et al. 2018) .- lambda
Strictly positive numeric value specifying the rate parameter of the gamma prior distribution of \(\alpha\). Defaults to
0.001. Whenn_cluster > 1, each mixture component \(\alpha_{c}\) has the same prior distribution.- psi
Positive integer specifying the concentration parameter \(\psi\) of the Dirichlet prior distribution used for the cluster probabilities \(\tau_{1}, \tau_{2}, \dots, \tau_{C}\), where \(C\) is the value of
n_clusters. Defaults to10L. Whenn_clusters = 1, this argument is not used.- kappa
Hyperparameters of the truncated Beta prior used for error probability \(\theta\) in the Bernoulli error model. The prior has the form \(\pi(\theta) = \theta^{\kappa_{1}} (1 - \theta)^{\kappa_{2}}\). Defaults to
c(1, 3), which means that the \(\theta\) is a priori expected to be closer to zero than to 0.5. See (Crispino et al. 2019) for details.
Value
An object of class "BayesMallowsPriors", to be provided in the
priors argument to compute_mallows(), compute_mallows_mixtures(), or
update_mallows().
References
Crispino M, Arjas E, Vitelli V, Barrett N, Frigessi A (2019).
“A Bayesian Mallows approach to nontransitive pair comparison data: How human are sounds?”
The Annals of Applied Statistics, 13(1), 492–519.
doi:10.1214/18-aoas1203
.
Vitelli V, Sørensen, Crispino M, Arjas E, Frigessi A (2018).
“Probabilistic Preference Learning with the Mallows Rank Model.”
Journal of Machine Learning Research, 18(1), 1–49.
https://jmlr.org/papers/v18/15-481.html.
See also
Other preprocessing:
get_transitive_closure(),
set_compute_options(),
set_initial_values(),
set_model_options(),
set_progress_report(),
set_smc_options(),
setup_rank_data()