| Project/Area Number |
23650142
|
|---|
| Research Category |
Grant-in-Aid for Challenging Exploratory Research
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | University of Tsukuba |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
YATA Kazuyoshi 筑波大学, 数理物質系, 助教 (90585803)
AKAHIRA Masafumi 筑波大学, 名誉教授 (70017424)
|
|---|
| Project Period (FY) |
2011 – 2013
|
| Project Status |
Completed (Fiscal Year 2013)
|
| Budget Amount *help |
¥3,380,000 (Direct Cost: ¥2,600,000、Indirect Cost: ¥780,000)
Fiscal Year 2013: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2012: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2011: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
|
|---|
| Keywords | 異常値 / クラスター分析 / 判別分析 / 多変量解析 / クロスデータ行列法 / 高次モーメント / 非ガウス / 回帰分析 / 逆行列 / 情報量規準 |
|---|
| Research Abstract |
In this research project, we aim to develop new multivariate statistical methods satisfying the criteria of high speed, robustness and high accuracy for inferences on modern data. We provided three multivariate statistical methods to ensure robustness and high accuracy with low computational cost even for non-Gaussian, contaminated models. The findings of this research are as follows: (1) Developments of high-speed and highly accurate classification methods using higher moments. (2) Developments of high-speed and highly accurate variable selection and outlier detection methods. (3) Intrinsic space analysis in a contaminated data space.
|
