Sparse variational approximations for partially confirmatory factor analysis

• In Event: Electronic Board Session III

Abstract

Different to MCMC-based methods for partially confirmatory factor analysis models (PCFA), this study proposes using sparse variational approximation approach as a compelling and faster alternative, which can achieve considerable accuracy in terms of computationally efficiency and accuracy-efficiency trade-offs, with scalability to large-scale problems.

Authors

• YI JIN

• Jinsong Chen, The University of Hong Kong