Model Selection for Singular Statistical Models and its Application
| Project/Area Number |
20700252
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | Kyushu University |
|---|
Principal Investigator |
NINOMIYA Yoshiyuki Kyushu University, 大学院・数理学研究院, 准教授 (50343330)
|
|---|
| Project Period (FY) |
2008 – 2010
|
| Project Status |
Completed (Fiscal Year 2010)
|
| Budget Amount *help |
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2010: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2009: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2008: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
|
|---|
| Keywords | 計量経済学 / 計量心理学 / 情報量規準 / 漸近分布論 / 統計幾何学 / 非正則性 / モデル選択 / 尤度比 / 漸近分布理論 / バイアス補正 / 非正則統計モデル / ブラウン運動 / 変化点モデル / モデル誤特定 / 疫学統計 / 集積性検出 / 信号検出 / 多重検定 / 特異モデル / 微分幾何 / 尤度比検定 / 因子分析モデル / 局所錐母数化 / モデル撰択 |
|---|
| Research Abstract |
Model selection is an indispensable task in statistical analysis. It is a method to select the optimal model from data in several model candidates. To use AIC is one of the most basic ideas for such model selection. The AIC is an estimator of the distance between the true distribution and the estimated distribution in a model candidate, and so the model which gives the minimum AIC is regarded as the optimal model. For so-called singular statistical models, however, any reasonable AIC has not derived until now. The achievement of this research is to derive a reasonable AIC for a structural change model and a factor analysis model, which are examples of the singular statistical models.
|
Report
(4 results)
Research Products
(17 results)
