Robust nonparametric inference for infinite variance processes by self-weighted empirical likelihood method
| Project/Area Number |
16K16022
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | The University of Tokyo (2019)
Waseda University (2016-2018) |
|---|
Principal Investigator |
Akashi Fumiya 東京大学, 大学院経済学研究科(経済学部), 講師 (90773268)
|
|---|
| Project Period (FY) |
2016-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000)
Fiscal Year 2019: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2018: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2017: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2016: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
|
|---|
| Keywords | 無限分散過程 / 一般化経験尤度法 / 自己加重法 / 自己基準化法 / 分位点回帰 / 時系列解析 / 統計的仮説検定 / 一般化経験尤度 / 無限分散時系列モデル / 長期記憶時系列モデル / 数理統計学 |
|---|
| Outline of Final Research Achievements |
This research project constructed robust statistical methods for infinite variance time series models and applied the methods to real data. In particular, the innovation process of the model can be infinite variance random variables, and we constructed natural likelihood-based inference procedures by using empirical likelihood approach. Moreover, the robust generalized empirical likelihood statistic is proposed by self-normalization and self-weighting methods. We also applied the fundamental methods to important practical problems including change point detection of the process, inference for time series regression models with long-memory disturbance, robust causality test of infinite variance models and model diagnostics. As a result, robust and novel statistical inference procedures for various nonregular processes are constructed.
|
| Academic Significance and Societal Importance of the Research Achievements |
近年、計量経済学・金融データ解析の分野では正規分布よりも裾の分布の厚い時系列データが観測され、古典的な尤度法やモーメント法を直接用いることができない。そこで本研究課題では有限・無限分散モデルを含む一般的なモデルに対して、モデル誤差項の分布を限定せずに解析を行い、統計量頑健化の手法により興味のない局外変数の事前推定が不要な統計手法を構成した。本研究課題で構成した手法では煩雑なチューニングパラメータの調整の必要はなく、局外母数の事前推定も不要である。結果として研究成果の概要の項目で述べた幅広い応用問題の枠組みでも頑健な統計手法を構成できることが示された。
|
Report
(5 results)
Research Products
(50 results)
