Σύγκριση μεθόδων

	Ανάλυση Λανθανουσών Κλάσεων (LCA)	Διακριτική Ανάλυση
Πεδίο	Στατιστική	Στατιστική
Οικογένεια	Latent structure	Latent structure
Έτος προέλευσης	1950s–1968	1936
Δημιουργός	Paul F. Lazarsfeld	Ronald A. Fisher
Τύπος	Latent variable / person-centered classification	Supervised classification and dimension reduction
Θεμελιώδης πηγή	Goodman, L. A. (1974). Exploratory latent structure analysis using both identifiable and unidentifiable models. Biometrika, 61(2), 215–231.	Fisher, R. A. (1936). The use of multiple measurements in taxonomic problems. Annals of Eugenics, 7(2), 179–188.
Εναλλακτικές ονομασίες	LCA, latent class model, latent categorical analysis, finite mixture of multinomials	LDA, Fisher discriminant analysis, discriminant function analysis, canonical discriminant analysis
Συναφείς	6	4
Σύνοψη	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.	Discriminant analysis finds linear combinations of predictor variables that best separate two or more known groups. It is used both to understand which predictors distinguish the groups and to classify new observations into those groups with minimum error.