WEKO3
アイテム
情報量規準TICとc-AICによるモデル選択の有効性
https://fra.repo.nii.ac.jp/records/2002167
https://fra.repo.nii.ac.jp/records/20021672745e337-3418-40ff-b2af-076b6172e7b1
| アイテムタイプ | 紀要論文 / Departmental Bulletin Paper(1) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 公開日 | 2024-04-22 | |||||||||||
| タイトル | ||||||||||||
| タイトル | 情報量規準TICとc-AICによるモデル選択の有効性 | |||||||||||
| 言語 | ja | |||||||||||
| タイトル | ||||||||||||
| タイトル | Efficiency of Model Selection by Information Criteria, TIC and c-AIC | |||||||||||
| 言語 | en | |||||||||||
| 言語 | ||||||||||||
| 言語 | jpn | |||||||||||
| キーワード | ||||||||||||
| 言語 | en | |||||||||||
| 主題Scheme | Other | |||||||||||
| 主題 | TIC; AIC; c-AIC; BIC; nested model; CPUE standardization | |||||||||||
| 資源タイプ | ||||||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||||||||
| 資源タイプ | departmental bulletin paper | |||||||||||
| アクセス権 | ||||||||||||
| アクセス権 | metadata only access | |||||||||||
| アクセス権URI | http://purl.org/coar/access_right/c_14cb | |||||||||||
| 著者 |
庄野, 宏
× 庄野, 宏
|
|||||||||||
| 抄録 | ||||||||||||
| 内容記述タイプ | Abstract | |||||||||||
| 内容記述 | In a model selection, AIC tends to overestimate the number of unknown parameters in several cases. However, c-AIC and T C correct the bias of AIC in small samples and where the true model does not include the candidate one in nested model, respectively. I calculated the values of AIC, c-AIC, TIC and BIC using a virtual example dealing with CPUE analysis. A finally selected model by c-AIC is simpler than that by other criteria (AIC, TIC and BIC). I carried out the computer simulation by ANOVA type model corresponding to CPUE standardization. Especially, I compared the selection performance of TIC to that of AIC c-AIC, and BIC using the nested model that the true model does not include candidate one. As a result of simulation, the efficiency of TIC is slightly better than that of AIC and c-AIC. In addition, the value of TIC is not so greatly different from that of AIC in the statistical model with normal error (such as ANOVA or linear regression), theoretically. Therefore, there seems no need to use TIC instead of AIC in such cases because of complexity of TIC formula. | |||||||||||
| 言語 | en | |||||||||||
| 書誌情報 |
ja : 遠洋水産研究所研究報告 en : Bulletin of National Research Institute of Far Seas Fisheries 巻 38, p. 21-28, ページ数 8, 発行日 2001-03 |
|||||||||||
| 出版者 | ||||||||||||
| 出版者 | 遠洋水産研究所 | |||||||||||
| 言語 | ja | |||||||||||
| 出版者 | ||||||||||||
| 出版者 | The National Research Institute of Far Seas Fisheries | |||||||||||
| 言語 | en | |||||||||||
| ISSN | ||||||||||||
| 収録物識別子タイプ | PISSN | |||||||||||
| 収録物識別子 | 0386-7285 | |||||||||||
| 書誌レコードID | ||||||||||||
| 収録物識別子タイプ | NCID | |||||||||||
| 収録物識別子 | AN00025949 | |||||||||||
| 情報源 | ||||||||||||
| 識別子タイプ | Local | |||||||||||
| 関連識別子 | enyo_k_38-21 | |||||||||||
| 関連サイト | ||||||||||||
| 識別子タイプ | URI | |||||||||||
| 関連識別子 | https://agriknowledge.affrc.go.jp/RN/2010622402 | |||||||||||
| 言語 | ja | |||||||||||
| 関連名称 | 日本農学文献記事索引 （AgriKnowledge） | |||||||||||
