Generalized Measurement Models
Abstract
Given a set of random variables, it is often the case that their associations can be explained by hidden common causes. The authors present a set of well-defined assumptions and a provably correct algorithm that allow them to identify some of those hidden common causes. The assumptions are fairly general and sometimes weaker than those used in practice by econometricians, psychometricians, social scientists, and by experts in many other fields in which latent variable models are important and tools such as factor analysis are applicable. The goal is automated knowledge discovery: identifying latent variables that can be used across different applications and causal models and that provide new insights into a data generating process. Their approach is evaluated through simulations and three real-world cases.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 27, 2005
- Accession Number
- ADA456031
Entities
People
- Ricardo S Silva
- Richard Scheines
Organizations
- Carnegie Mellon University