Theory and Applications of Density Ratio
| Project/Area Number |
24500340
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | Nagoya University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2012-04-01 – 2016-03-31
|
| Project Status |
Completed (Fiscal Year 2015)
|
| Budget Amount *help |
¥5,330,000 (Direct Cost: ¥4,100,000、Indirect Cost: ¥1,230,000)
Fiscal Year 2014: ¥2,730,000 (Direct Cost: ¥2,100,000、Indirect Cost: ¥630,000)
Fiscal Year 2013: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2012: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
|
|---|
| Keywords | 数理統計学 / 機械学習 / 統計学 / 統計的学習理論 / ロバスト統計 |
|---|
| Outline of Final Research Achievements |
Density ratio is defined as the ratio of two probability densities. That is very useful when two distinct data domains are compared. Important applications of the density ratio includes regression analysis and classification problems under covariate shift, statistical test, and dimension reduction. In this study, we developed methods to improve the statistical accuracy and computational efficiency of density ratio estimation. Moreover, we studied the relation between density ration and statistical divergences which is regarded as a discrepancy measure between two probability distributions. Statistical usefulness of the divergence is considered from the view point of the density ratio.
|
Report
(5 results)
Research Products
(50 results)
-
-
-
-
-
-
-
-
-
-
[Journal Article] Direct Divergence Approximation between Probability Distributions and Its Applications in Machine Learning2013
Author(s)
Sugiyama, M., Liu, S., du Plessis, M. C., Yamanaka, M., Yamad a, M., Suzuki, T., & Kanamori, T.
-
Journal Title
Journal of Computing Science and Engineering
Volume: Vol.7no.2
Pages: 99-111
DOI
Related Report
Peer Reviewed
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
