Latent and observable variables: Difference between revisions

Revision as of 13:05, 24 October 2006

Revision as of 08:30, 31 October 2006

Line 1:

'''Latent variables''', as opposed to [[observable variable]]s, are [[variable]]s that are not directly observed but are rather inferred from other variables that are observed and directly measured. Latent variables are also called '''hidden variables''', '''model parameters''', '''hypothetical variables''' or '''hypothetical constructs'''. Examples include [[quality of life]], business confidence, morale, happiness, conservatism. The use of latent variables is common in [[social science]]s, [[robotics]], and to an extent [[economics]]; but the exact definition of latent variables varies in these domains. [[John F. MacGregor]] at [[McMaster University]] pioneered their use in [[chemical engineering]] and [[control systems]] by using them as controlled variables in [[model predictive control]].

One advantage of using latent variables is that it [[Dimensionality reduction|reduces the dimensionality]] of data. A large number of observable variables can be aggregated to represent an underlying concept, making it easier for humans to understand the data. In this sense, they serve the same function as theories in general do in science. At the same time, latent variables are link observable ("sub-symbolic") data in the real world, to symbolic data in the modeled world.

==See also==