Prediction and Conditional Normalized Maximum Likelihood Distribution
| Project/Area Number |
26730014
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | Hokkaido University (2017)
The University of Tokyo (2014-2016) |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2014-04-01 – 2018-03-31
|
| Project Status |
Completed (Fiscal Year 2017)
|
| Budget Amount *help |
¥3,770,000 (Direct Cost: ¥2,900,000、Indirect Cost: ¥870,000)
Fiscal Year 2017: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2016: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2015: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2014: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
|
|---|
| Keywords | 統計的予測 / 統計的決定理論 / 予測 / 条件付き正規化最尤分布 / 情報幾何 |
|---|
| Outline of Final Research Achievements |
This research project treated the conditional normalized maximum likelihood distribution, which is a conditional probability distribution. We were also interested in a measure of probability distributions which is called the conditional regret. When we fix past observations and when we consider the average of the past observations, another probability distribution is superior to the conditional normalized maximum likelihood distribution if they are evaluated with the Kullback-Leibler information. That probability distribution attains the minimax of another measure of probability distributions, which is called the conditional regret risk. Furthermore, the probability distribution coincides with the projection of the conditional normalized maximum likelihood distribution on the space of the Bayesian predictions.
|
Report
(5 results)
Research Products
(13 results)
