Parameter estimation and uncertainty

The chapter contains examples of procedures for estimation of parameters in probability models for relationship inference. The main examples are the theta correction parameter, the lambda parameter for haplotype estimation, and the mutation rate parameter. The same general Bayesian framework is applied to all these, and can be extended to other parameters. It is also discussed how the likelihoods of likelihood ratios should be computed by integrating out uncertainty in uncertain parameters. One should avoid talking about uncertainty in likelihood ratios, but should instead talk about how likelihood ratios might change when distributions of uncertain parameters change.