1. Introduction -- 2. State-Space Model in Linear Case -- 3. Traditional Nonlinear Filters -- 4. Density-Based Nonlinear Filters -- 5. Monte-Carlo Experiments -- 6. Application of Nonlinear Filters -- 7. Prediction and Smoothing -- 8. Summary and Concluding Remarks -- References.

Nonlinear and nonnormal filters are introduced and developed. Traditional nonlinear filters such as the extended Kalman filter and the Gaussian sum filter give biased filtering estimates, and therefore several nonlinear and nonnormal filters have been derived from the underlying probability density functions. The density-based nonlinear filters introduced in this book utilize numerical integration, Monte-Carlo integration with importance sampling or rejection sampling and the obtained filtering estimates are asymptotically unbiased and efficient. By Monte-Carlo simulation studies, all the nonlinear filters are compared. Finally, as an empirical application, consumption functions based on the rational expectation model are estimated for the nonlinear filters, where US, UK and Japan economies are compared.