Salīdzināt metodes

	Multidimensionālā skalēšana (MDS)	Latent Class Analysis (LCA)
Nozare	Statistika	Statistika
Saime	Latent structure	Latent structure
Izcelsmes gads	1952–1964	1950s–1968
Autors	Warren S. Torgerson (metric MDS, 1952); Joseph B. Kruskal (non-metric MDS, 1964)	Paul F. Lazarsfeld
Tips	Dimensionality reduction / visualization	Latent variable / person-centered classification
Pirmavots	Kruskal, J. B. (1964). Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika, 29(1), 1–27.	Goodman, L. A. (1974). Exploratory latent structure analysis using both identifiable and unidentifiable models. Biometrika, 61(2), 215–231.
Citi nosaukumi	MDS, metric MDS, non-metric MDS, proximity scaling	LCA, latent class model, latent categorical analysis, finite mixture of multinomials
Saistītās	5	6
Kopsavilkums	Multidimensional scaling maps objects described only by pairwise similarities or dissimilarities into a low-dimensional geometric space so that distances in that space reflect the original proximity structure as faithfully as possible. It is widely used to visualize the hidden structure of psychological, social, and behavioral data.	Latent class analysis identifies unobserved subgroups — latent classes — within a population by finding patterns of responses across a set of categorical observed indicators. It is the categorical-variable counterpart of cluster analysis, but grounded in an explicit probabilistic model, and is widely used in social, health, and behavioral sciences to discover typologies in survey or diagnostic data.