Dissertation > Excellent graduate degree dissertation topics show

The Simulation of Bayesian Dynamic Models

Author: SuBing
Tutor: JiangXiaoZuo
School: Shandong University
Course: Applied Mathematics
Keywords: Bayesian dynamic model Stochastic simulation Bayes factor Model monitoring
CLC: O212.8
Type: Master's thesis
Year: 2007
Downloads: 94
Quote: 1
Read: Download Dissertation


This paper studies non-the normality assumption dynamic linear model observation equation: y t = F t T θ t ν t , ν t ~ [0, V t ]; equation of state: θ t = G t θ t-1 ω t , ω t ~ [0, W t ]. (\t ) ~ p (y t | θ t ); state equation: (y t | θ t- 1 ) to P (y t | θ T-1 ) (\function) simulation algorithm, and given a new Bayesian model, and finally discuss the Bayesian model selection and monitoring, which reads as follows: Bayesian theory of the development of the process outlined in the first chapter, introduced Tony dynamic model theory Julius basic concepts and problems encountered in the current research in this area, and the framework of the entire paper with a brief summary. stochastic simulation methods introduced in the second chapter, the main importance of sampling Metropolis-Hastings sampling, Gibbs sampling, three related to this article. latter two methods occupies an important position in the dynamic model simulations, we have carried out a more detailed study, including sample extraction, sample convergence diagnostic specifically given a diagnosis method: Riemann and convergence diagnostics in the third chapter, stochastic simulation methods to deal with these two models above the main idea: instead of the density function of the sample correction recursive recursive correction, and then the samples for a variety of inference; different recursive algorithm for a different model, in this process, in this chapter, a number of convex density new dynamic model, and prove that it meets the \Luo ways to strike a Bayes factor, then through Bayes factor model selection. the Bayes factor mentioned in the literature [2] in the complex case is no longer applicable, proposed a new Bayes factor, it will be discussing the model is extremely useful given two model monitoring methods in Chapter One is the use of the observation error constructed a χ 2 - Statistics amount, using the statistics to the accuracy of the set is given a truncation point with the truncation point on the model to achieve the local and overall monitoring; second is to propose a new Bayesian factor, which is adjacent than the moment observations marginal density, can achieve a partial monitoring of the model this moment.

Related Dissertations

  1. Modeling and Analysis of Asset-liability Management for Life Insurance Companies of China,F842.3
  2. Research on Stochastic Simulation Approaches for Spatially Inhomogeneous Chemical Reaction Systems Based on PDES,TQ019
  3. The Stochastic Simulation Method Application in Design Flood of Manas River Basin,TV122.3
  4. The Study Andapplication of the Multiscale Seismic Data Joint Inversion Method,P618.13
  5. Study on the Comprehensive Optimization of Train Formation Plan and Flow Path Based on Uncertain Parameters,U292.31
  6. Coordinated Strategy Research Based on Manufacturing/Remanufacturing Closed Loop Supply Chain,F274
  7. Medical Monitoring Model Based on Wireless Sensor Network,TN929.5;TP277
  8. Stochastic Simulation Based on FAHP quality with cost-cutting Strategies,F273.2
  9. Multi-span continuous beam bridge seismic response of multi-point non- uniform excitation,P315.9
  10. Geological Modeling on the Reservoir of Complicated Fault Block,P618.13
  11. Statistical methods in calculating resource utilization of flood risk management in applied research,TV213.9
  12. Caving ore dimensional numerical simulation of random media,TD853.36
  13. Research on Energy-saving and Emission-reduction Evaluation of the Coal Resource-based Cities in China,F206
  14. Determination of Paleo-saltwater Intrusion Processes by Use of Deterministic and Stochastic Models,P641.2
  15. Xiamen Bay oil spill pollution accident hazard risk assessment studies,X55
  16. The Application of Lasso and Its Related Methods in Multiple Linear Regression Model,O212.1
  17. China's coal resources in the price formation mechanism,F426.21
  18. Application of Stochastic Simulation in 3D Geological Modeling,P618.13
  19. Analysis on Water-supply Risk and Reliability of Daduhe River Hydro-junction,TV674
  20. The Application of Dynamic Financial Analysis Model for the Non-life Insurance Company,F840
  21. Inversion Analysis of Tunnel Rock Parameter Based on Monitoring and Numerical Simulation,U452.12

CLC: > Mathematical sciences and chemical > Mathematics > Probability Theory and Mathematical Statistics > Mathematical Statistics > Bayesian statistics
© 2012 www.DissertationTopic.Net  Mobile