The msBayes model, and my modification of it, is a model-choice method with the primary purpose of estimating the probabilities of models of divergence across taxa.
An appealing approach would be a comparative, Bayesian model-choice method for inferring the probability of competing divergence histories while integrating over uncertainty in mutational and ancestral processes via models of nucleotide substitution and lineage coalescence.
Finally, as illustration, the timetabling MINLP model with both mode choice methods is applied to a real case and computational results are shown.
Comparison of models has been performed both by the model choice acceptance method and by using Bayes factors between each pair of models, two different ways leading to identical conclusions.
To evaluate whether the observed data were in agreement with the simulated data, ABCtoolbox reports a p-value varying between 0 (no fit) and 1 (good fit) [ 33].> The relative probabilities of the three scenarios were computed using the model choice acceptance method [ 42] consisting in the proportion δ of data sets, among the N simulated, that are the closest to the observed data set.
We suggest employing two commonly used model selection methods, namely MDL and CV, for making such choices.
Multinomial probit model may be a preferred choice method of analysis if we assume the existence of the correlation between alternatives.
Only after careful model choice and optimization can the method be applied routinely and data analyzed automatically by batches.
These parameters were then estimated using our method, with model choice informed by the rescaled deviance information criterion.
This aspect of weighting of information is very closely linked to Section "Model criticism", as the correct assumption can be examined through methods for model choice.
Here, we present an example where the decision on how to treat sites with missing data is of equal importance to decisions on taxon sampling and model choice, and we introduce a graphical method for illustrating this.
