作者: Kohei ADACHI
DOI: 10.2333/JBHMK.27.12
关键词:
摘要: A metric multidimensional unfolding procedure is proposed to represent the rows (individuals) and columns (items) of a proximity data matrix as points in low-dimensional space. The based on random effect model which allows us avoid problem incidental parameters. In model, individual are regarded normally-distributed variables, while item fixed probability density derived from assumption that true linear function distance point observed perturbed by error. marginal likelihood obtained integrating out maximized using EM algorithm with generalized SMACOF algorithm. evaluated simulation study applied preference rating data.