WEAK PROBABILITY MODELLING WITH APPLICATION TO DYNAMIC MODELS By Kostas Triantafyllopoulos and Jeff Harrison University of Bristol and University of Warwick
暂无分享,去创建一个
Partially specified probability models are constructed based upon korder conditional independence, in order to facilitate updating and prediction of required distributional quantities. Weak models relate to second order conditional independence and are formulated particularly for multivariate dynamic linear models with unknown observational variance matrices. The resulting sequential variance updating algorithm is applied to a simulated series in order both to judge its efficiency and to compare it with an existing method.
[1] J. Hartigan. Linear Bayesian Methods , 1969 .
[2] Jeff Harrison,et al. Subjective intervention in formal models , 1989 .
[3] D. Harville. Matrix Algebra From a Statistician's Perspective , 1998 .
[4] F P Wheeler,et al. Bayesian Forecasting and Dynamic Models (2nd edn) , 1998, J. Oper. Res. Soc..
[5] Ute Beyer,et al. Bayesian Forecasting And Dynamic Models , 2016 .