EM-algorithm;
Grobner basis;
identifiability;
maximum likelihood;
network tomography;
sufficient statistic;
D O I:
10.1023/A:1020775726595
中图分类号:
TP301 [理论、方法];
学科分类号:
081202 ;
摘要:
We describe applications of computational algebra to statistical problems of parameter identifiability, sufficiency, and estimation. The methods work for a family of statistical models that includes Poisson and binomial examples in network tomography.
机构:
Cedars Sinai Med Ctr, Biostat & Bioinformat Res Ctr, Los Angeles, CA 90048 USACedars Sinai Med Ctr, Biostat & Bioinformat Res Ctr, Los Angeles, CA 90048 USA
机构:
Univ Calif Los Angeles, Dept Epidemiol, Los Angeles, CA USA
Univ Calif Los Angeles, Dept Stat, Los Angeles, CA USAUniv Calif Los Angeles, Dept Epidemiol, Los Angeles, CA USA