Algebraic Sparse Factor Analysis

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Drton, Mathias, Grosdos, Alexandros, Portakal, Irem, Sturma, Nils
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912146898550784
author Drton, Mathias
Grosdos, Alexandros
Portakal, Irem
Sturma, Nils
author_facet Drton, Mathias
Grosdos, Alexandros
Portakal, Irem
Sturma, Nils
contents Factor analysis is a statistical technique that explains correlations among observed random variables with the help of a smaller number of unobserved factors. In traditional full factor analysis, each observed variable is influenced by every factor. However, many applications exhibit interesting sparsity patterns, that is, each observed variable only depends on a subset of the factors. In this paper, we study such sparse factor analysis models from an algebro-geometric perspective. Under mild conditions on the sparsity pattern, we examine the dimension of the set of covariance matrices that corresponds to a given model. Moreover, we study algebraic relations among the covariances in sparse two-factor models. In particular, we identify cases in which a Gröbner basis for these relations can be derived via a 2-delightful term order and join of toric ideals of graphs.
format Preprint
id arxiv_https___arxiv_org_abs_2312_14762
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Algebraic Sparse Factor Analysis
Drton, Mathias
Grosdos, Alexandros
Portakal, Irem
Sturma, Nils
Statistics Theory
Commutative Algebra
62H25, 62R01, 13F65, 14M25, 14N07
Factor analysis is a statistical technique that explains correlations among observed random variables with the help of a smaller number of unobserved factors. In traditional full factor analysis, each observed variable is influenced by every factor. However, many applications exhibit interesting sparsity patterns, that is, each observed variable only depends on a subset of the factors. In this paper, we study such sparse factor analysis models from an algebro-geometric perspective. Under mild conditions on the sparsity pattern, we examine the dimension of the set of covariance matrices that corresponds to a given model. Moreover, we study algebraic relations among the covariances in sparse two-factor models. In particular, we identify cases in which a Gröbner basis for these relations can be derived via a 2-delightful term order and join of toric ideals of graphs.
title Algebraic Sparse Factor Analysis
topic Statistics Theory
Commutative Algebra
62H25, 62R01, 13F65, 14M25, 14N07
url https://arxiv.org/abs/2312.14762