Generalized Monotone Additive Latent Variable Models
affiliation not provided to SSRN
University of Geneva - HEC
May 4, 2010
For manifest variables with additive noise and for a given number of latent variables with an assumed distribution, we propose to nonparametrically estimate the association between latent and manifest variables. Our estimation is a two step procedure: first it employs standard factor analysis to estimate the latent variables as theoretical quantiles of the assumed distribution; second, it employs the additive models’ backfitting procedure to estimate the monotone nonlinear associations between latent and manifest variables. The estimated fit may suggest a different latent distribution or point to nonlinear associations. We show on simulated data how, based on mean squared errors, the nonparametric estimation improves on factor analysis. We then employ the new estimator on real data to illustrate its use for exploratory data analysis.
Number of Pages in PDF File: 24
Keywords: Factor Analysis, Principal Component Analysis, Nonparametric Regression, Bartlett’s Factor Scores, Dimension Reductionworking papers series
Date posted: February 17, 2011
© 2013 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo6 in 0.359 seconds