4. Conclusion

We refer you to for much more detailed examples in the context of generalized linear models, capture-recapture models, mixture models, time series... In each case, there is a default a priori model and an algorithmic resolution that provides a Bayesian reference solution for the problem under consideration. Of course, other a priori laws can be considered, and the reference model is then used to evaluate the impact of this a priori choice. We simply wanted to convey the idea that it is possible to carry out Bayesian inference on a realistic problem without any particular expertise in the construction of a priori laws.