作者: Shingo Saburi , Naohito Chino
DOI:
关键词: Restricted maximum likelihood 、 Statistics 、 Likelihood-ratio test 、 Multidimensional scaling 、 Score test 、 Statistic 、 Context (language use) 、 Mathematics 、 Maximum likelihood sequence estimation 、 Likelihood function
摘要: ASYMMAXSCAL is revisited first, which a maximum likelihood asymmetric multidimensional scaling method recently proposed by Saburi and Chino (2008). It proven that the ratio test statistic on quasi-symmetry hypothesis Caussinus (1965) of marginal homogeneity suggested Andersen (1980) are mutually independent statistically. A possible application this theorem indicated to relational data in context scaling.