| Project/Area Number |
24700284
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | Kansai University |
|---|
Principal Investigator |
TAKAI Keiji 関西大学, 商学部, 准教授 (20572019)
|
|---|
| Project Period (FY) |
2012-04-01 – 2014-03-31
|
| Project Status |
Completed (Fiscal Year 2013)
|
| Budget Amount *help |
¥2,860,000 (Direct Cost: ¥2,200,000、Indirect Cost: ¥660,000)
Fiscal Year 2013: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2012: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
|
|---|
| Keywords | 欠測データ / 最尤法 / 漸近理論 / 判別分析 / 部分的にラベルづけされたデータ / 無視可能な欠測 / 欠測データメカニズム / 部分的にラベル付けされたデータ / 欠測値 / 多変量解析 / 無視不可能な欠測 / 推定関数 |
|---|
| Research Abstract |
As the first result of my study, I found that the maximum likelihood estimator (MLE) constructed from incomplete data has desirable properties such as consistency and asymptotic normality under the different conditions to the case in which completely observed data are available. The second result is application of the first result to discriminant analysis with partially labeled data. Since the partially labeled data can be regarded as missing data, the first result can also be applied to estimation of the parameters in such discriminant model when constructing a discriminant rule. It is found that all data available to us should be used when the observations used to construct the rule is completely randomly chosen, while there are times when not all observations with or without labels should be used for the data which are chosen depending on the value of the feature vector.
|
