ログイン
言語:

WEKO3

  • トップ
  • ランキング
To
lat lon distance
To

Field does not validate



インデックスリンク

インデックスツリー

メールアドレスを入力してください。

WEKO

One fine body…

WEKO

One fine body…

アイテム

{"_buckets": {"deposit": "7cdcda12-f7c0-4d33-bdd6-b3fdc085b6b8"}, "_deposit": {"created_by": 18, "id": "2000323", "owner": "18", "owners": [18], "owners_ext": {"displayname": "NII", "username": "niirepo"}, "pid": {"revision_id": 0, "type": "depid", "value": "2000323"}, "status": "published"}, "_oai": {"id": "oai:repository.nii.ac.jp:02000323", "sets": ["136"]}, "author_link": [], "item_5_biblio_info_30": {"attribute_name": "書誌情報", "attribute_value_mlt": [{"bibliographicIssueDates": {"bibliographicIssueDate": "2014-09-25", "bibliographicIssueDateType": "Issued"}, "bibliographicPageEnd": "22", "bibliographicPageStart": "1", "bibliographic_titles": [{"bibliographic_title": "NIIテクニカル・レポート", "bibliographic_titleLang": "ja"}, {"bibliographic_title": "NII Technical Report", "bibliographic_titleLang": "en"}]}]}, "item_5_description_28": {"attribute_name": "抄録", "attribute_value_mlt": [{"subitem_description": "Nonlinear least squares (NLS) problems arise in many applications. The common solvers require to compute and store the corresponding Jacobian matrix explicitly, which is too expensive for large problems. In this paper, we propose an effective Jacobian free method especially for large NLS problems because of the novel combination of using automatic differentiation for J(x)v and JT (x)v along with the preconditioning ideas that do not require forming the Jacobian matrix J(x) explicitly. Together they yield a new and effective three-level iterative approach. In the outer level, the dogleg/trust region method is employed to solve the NLS problem. At each iteration of the dogleg method, we adopt the iterative linear least squares (LLS) solvers, CGLS or BA-GMRES method, to solve the LLS problem generated at each step of the dogleg method as the middle iteration. In order to accelerate the convergence of the iterative LLS solver, we propose an inner iteration preconditioner based on the weighted Jacobi method. Compared to the common dogleg solver and truncated Newton method, our proposed three level method need not compute the gradient or Jacobian matrix explicitly, and is efficient in computational complexity and memory storage. Furthermore, our method does not rely on the sparsity or structure pattern of the Jacobian, gradient or Hessian matrix. Thus, it can be applied to solve any large general NLS problem. Numerical experiments show that our proposed method is much superior to the common trust region method and truncated Newton method.", "subitem_description_language": "en", "subitem_description_type": "Abstract"}]}, "item_5_identifier_registration": {"attribute_name": "ID登録", "attribute_value_mlt": [{"subitem_identifier_reg_text": "10.20736/0002000323", "subitem_identifier_reg_type": "JaLC"}]}, "item_5_publisher_31": {"attribute_name": "出版者", "attribute_value_mlt": [{"subitem_publisher": "国立情報学研究所", "subitem_publisher_language": "ja"}]}, "item_5_source_id_32": {"attribute_name": "ISSN", "attribute_value_mlt": [{"subitem_source_identifier": "1346-5597", "subitem_source_identifier_type": "ISSN"}]}, "item_creator": {"attribute_name": "著者", "attribute_type": "creator", "attribute_value_mlt": [{"creatorNames": [{"creatorName": "Xu, Wei", "creatorNameLang": "en"}]}, {"creatorNames": [{"creatorName": "Zheng, Ning", "creatorNameLang": "en"}]}, {"creatorNames": [{"creatorName": "速水, 謙", "creatorNameLang": "ja"}, {"creatorName": "Hayami, Ken", "creatorNameLang": "en"}]}]}, "item_files": {"attribute_name": "ファイル情報", "attribute_type": "file", "attribute_value_mlt": [{"accessrole": "open_date", "date": [{"dateType": "Available", "dateValue": "2022-06-08"}], "displaytype": "detail", "download_preview_message": "", "file_order": 0, "filename": "14-003E.pdf", "filesize": [{"value": "304 KB"}], "format": "application/pdf", "future_date_message": "", "is_thumbnail": false, "licensetype": "license_free", "mimetype": "application/pdf", "size": 304000.0, "url": {"label": "NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems", "url": "https://repository.nii.ac.jp/record/2000323/files/14-003E.pdf"}, "version_id": "e19d12a8-5381-46f6-bd83-6edeedc787f4"}]}, "item_keyword": {"attribute_name": "キーワード", "attribute_value_mlt": [{"subitem_subject": "テクニカルレポート", "subitem_subject_language": "ja", "subitem_subject_scheme": "Other"}, {"subitem_subject": "Technical Report", "subitem_subject_language": "en", "subitem_subject_scheme": "Other"}]}, "item_language": {"attribute_name": "言語", "attribute_value_mlt": [{"subitem_language": "eng"}]}, "item_resource_type": {"attribute_name": "資源タイプ", "attribute_value_mlt": [{"resourcetype": "departmental bulletin paper", "resourceuri": "http://purl.org/coar/resource_type/c_6501"}]}, "item_title": "NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems", "item_titles": {"attribute_name": "タイトル", "attribute_value_mlt": [{"subitem_title": "NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems", "subitem_title_language": "en"}]}, "item_type_id": "5", "owner": "18", "path": ["136"], "permalink_uri": "https://doi.org/10.20736/0002000323", "pubdate": {"attribute_name": "PubDate", "attribute_value": "2022-06-08"}, "publish_date": "2022-06-08", "publish_status": "0", "recid": "2000323", "relation": {}, "relation_version_is_last": true, "title": ["NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems"], "weko_shared_id": -1}
  1. NIIテクニカル・レポート

NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems

https://doi.org/10.20736/0002000323
https://doi.org/10.20736/0002000323
1deb57c4-ab35-4912-b069-a31e1f127011
名前 / ファイル ライセンス アクション
14-003E.pdf NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems (304 KB)
Item type レポート / Report(1)
公開日 2022-06-08
タイトル
言語 en
タイトル NII Technical Report (NII-2014-003E):Jacobian-Free Three-Level Trust Region Method for Nonlinear Least Squares Problems
言語
言語 eng
キーワード
言語 ja
主題Scheme Other
主題 テクニカルレポート
キーワード
言語 en
主題Scheme Other
主題 Technical Report
資源タイプ
資源 http://purl.org/coar/resource_type/c_6501
タイプ departmental bulletin paper
ID登録
ID登録 10.20736/0002000323
ID登録タイプ JaLC
著者 Xu, Wei

× Xu, Wei

en Xu, Wei

Search repository
Zheng, Ning

× Zheng, Ning

en Zheng, Ning

Search repository
速水, 謙

× 速水, 謙

ja 速水, 謙

en Hayami, Ken

Search repository
抄録
内容記述タイプ Abstract
内容記述 Nonlinear least squares (NLS) problems arise in many applications. The common solvers require to compute and store the corresponding Jacobian matrix explicitly, which is too expensive for large problems. In this paper, we propose an effective Jacobian free method especially for large NLS problems because of the novel combination of using automatic differentiation for J(x)v and JT (x)v along with the preconditioning ideas that do not require forming the Jacobian matrix J(x) explicitly. Together they yield a new and effective three-level iterative approach. In the outer level, the dogleg/trust region method is employed to solve the NLS problem. At each iteration of the dogleg method, we adopt the iterative linear least squares (LLS) solvers, CGLS or BA-GMRES method, to solve the LLS problem generated at each step of the dogleg method as the middle iteration. In order to accelerate the convergence of the iterative LLS solver, we propose an inner iteration preconditioner based on the weighted Jacobi method. Compared to the common dogleg solver and truncated Newton method, our proposed three level method need not compute the gradient or Jacobian matrix explicitly, and is efficient in computational complexity and memory storage. Furthermore, our method does not rely on the sparsity or structure pattern of the Jacobian, gradient or Hessian matrix. Thus, it can be applied to solve any large general NLS problem. Numerical experiments show that our proposed method is much superior to the common trust region method and truncated Newton method.
言語 en
書誌情報 ja : NIIテクニカル・レポート
en : NII Technical Report

p. 1-22, 発行日 2014-09-25
出版者
言語 ja
出版者 国立情報学研究所
ISSN
収録物識別子タイプ ISSN
収録物識別子 1346-5597
戻る
0
views
See details
Views

Versions

Ver.1 2022-06-08 04:22:21.061531
Show All versions

Share

Mendeley Twitter Facebook Print Addthis

Cite as

エクスポート

OAI-PMH
  • OAI-PMH JPCOAR
  • OAI-PMH DublinCore
  • OAI-PMH DDI
Other Formats
  • JSON
  • BIBTEX

Confirm


Powered by WEKO3


Powered by WEKO3