Multivariate logistic compounds

摘要: SUMMARY A general method for estimating the proportions of a finite compound distribution is suggested, based on direct estimation likelihood ratio. The only distributional assumption that log ratio linear in observations and hence robust applicable to number families distributions. Sample points are required from each component compound. In particular, handles continuous discrete data with equal facility. Only moderate parameters need so problems quite high dimensions can be solved. used update logistic discriminant functions using whose parent populations unknown.