Equivalence-Motivated Non-Linear
Recursive Estimation}
Recursive non-linear Bayesian estimation is addressed using
equivalence approach as motivating framework. Its specific form
tailored to a model class covering non-normal ARX (auto-regression
with exogenous variables) models, logistic-type models and their
dynamic mixtures is presented. The resulting algorithms provide a
novel and efficient solutions of difficult and practically
important estimation problems.