This thesis is concerned with the approximations of prediction error and predictive distribution of the best reserve estimate produced by the models which combine information from multiple data sets. Two models are studied . ~; within the GLM framework, i.e. SClmieper's: model proposed by Schnieper (1991) and the MeL method introduced by Quargand Mack (2004). Theoretical and empirical approximation approaches for the MSEP of these two models are discussed and compared. This includes derivations of closed formulae following the approaches introduced by· both Mack (1993) and Murphy (1994) and also the empirical approach, i.e. the bootstrap method. And finally, various models which combine information from multiple data sets are investigated and compared, providing new insights to the claims .reserving area.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:492349 |
| Date | January 2008 |
| Creators | Liu, Huijuan |
| Publisher | City University London |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.002 seconds