Research On The Capture-Mark-Recapture Model

Capture-Recapture model is a classic method to solve the problem of group number est imat ing,being widely used in many fields about quant it y est imat ion.In this thesis,first ly we have a brief introduction about the procedure of this method,so me basic mode ls,and the research findings by R.M.Fewster and R.E.Jupp about the normal form of the model.Then we try to program to simulate to get the results,mainly for the maximum likelihood estimat io n and the condit io nal maximum likelihood estimat io n of the sampl e size.We also use the Bootstrap method and the Stochast ic perturbat ion met hod to get results and compare them wit h the result s fro m normal method.Finally the simulat io n result s verified so me conclusio ns from R.M.Fewster and R.E.Jupp,we find that these three methods have clo se result s when the model is quite simple,and the Stochast ic perturbat ion method get better results when the model is co mplex.