Estimation of High Dimensional Vector Autoregression via Sparse Precision Matrix

Poignard, Benjamin; Asai, Manabu

Estimation of High Dimensional Vector Autoregression via Sparse Precision Matrix

Poignard, Benjamin; Asai, Manabu

Discussion Papers In Economics And Business, 2021, 21-03, 1-66

アクセス数:268件（2026-04-05 06:29 集計）

固定URL: https://hdl.handle.net/11094/81251

論文情報

ファイル出力 EndNote Basic出力 Mendeley出力

タイトル	Estimation of High Dimensional Vector Autoregression via Sparse Precision Matrix
著者	NRID 1000040845252 1000040845252 ORCID 0000-0001-6110-6819 0000-0001-6110-6819 researchmap permalink benjamin.poignard benjamin.poignard Scopus Author ID 57201668513 57201668513 著者名 Poignard, Benjamin
著者	NRID 1000090319484 1000090319484 著者名 Asai, Manabu
キーワード等	sparse structural vector autoregression
	statistical consistency
	support recovery
抄録	We consider the problem of estimating sparse structural vector autoregression (SVAR) processes via penalized precision matrix. Such matrix is the output of the underlying directed acyclic graph of the SVAR process, whose zero components correspond to zero SVAR coe?cients. The precision matrix estimators are deduced from the class of Bregman divergences and regularized by the SCAD, MCP and LASSO penalties. Under suitable regularity conditions, we derive error bounds for the regularized precision matrix for each Bregman divergence. Moreover, we establish the support recovery property, including the case when the penalty is non-convex. These theoretical results are supported by empirical studies.
出版者	Graduate School of Economics, Osaka University
収録物名	Discussion Papers In Economics And Business
巻(号)	21-03
ページ	1-66
刊行年月	2021-04
言語	英語
ハンドルURL	https://hdl.handle.net/11094/81251
NCID	AA1160366X
関連情報 (references)	URI http://www2.econ.osaka-u.ac.jp/econ_society/dp/2103.pdf
アクセス権	metadata only access
カテゴリ	ディスカッション／ワーキング・ペーパー DP/WP
カテゴリ	Discussion Papers In Economics And Business

資源タイプ	テクニカルレポート
ローカル資源タイプ	ディスカッション／ワーキング・ペーパー
DCMI資源タイプ	text
DCTERMS.bibliographicCitation	Discussion Papers In Economics And Business. 2021, 21-03, p. 1-66
DC.title	Estimation of High Dimensional Vector Autoregression via Sparse Precision Matrix
DC.creator	Poignard, Benjamin
DC.creator	Asai, Manabu
DC.publisher	Graduate School of Economics, Osaka University
DC.language' scheme='DCTERMS.RFC1766	英語
DC.type' scheme='DCTERMS.DCMIType	text
DCTERMS.issued' scheme='DCTERMS.W3CDTF	2021-04
DC.relation' scheme='DCTERMS.URI	http://www2.econ.osaka-u.ac.jp/econ_society/dp/2103.pdf
DC.subject	sparse structural vector autoregression
	statistical consistency
	support recovery
DCTERMS.abstract	We consider the problem of estimating sparse structural vector autoregression (SVAR) processes via penalized precision matrix. Such matrix is the output of the underlying directed acyclic graph of the SVAR process, whose zero components correspond to zero SVAR coe?cients. The precision matrix estimators are deduced from the class of Bregman divergences and regularized by the SCAD, MCP and LASSO penalties. Under suitable regularity conditions, we derive error bounds for the regularized precision matrix for each Bregman divergence. Moreover, we establish the support recovery property, including the case when the penalty is non-convex. These theoretical results are supported by empirical studies.
DC.format	81251
citation_title	Estimation of High Dimensional Vector Autoregression via Sparse Precision Matrix
citation_author	Poignard, Benjamin
citation_author	Asai, Manabu
citation_publisher	Graduate School of Economics, Osaka University
citation_language	英語
citation_date	2021-04
citation_journal_title	Discussion Papers In Economics And Business
citation_volume	21-03
citation_firstpage	1
citation_lastpage	66
citation_public_url	https://hdl.handle.net/11094/81251
citation_abstract_html_url	http://www2.econ.osaka-u.ac.jp/econ_society/dp/2103.pdf

Estimation of High Dimensional Vector Autoregression via Sparse Precision Matrix

論文情報

論文詳細を表示