2011 Fiscal Year Final Research Report
Discovery of reliable causal structures in high-dimensional data
| Project/Area Number |
21700302
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | Osaka University |
|---|
Principal Investigator |
SHIMIZU Shohei 大阪大学, 産業科学研究所, 助教 (10509871)
|
|---|
| Project Period (FY) |
2009 – 2011
|
| Keywords | 統計的因果推論 / 構造方程式モデル / 独立成分分析 / 因果構造探索 |
|---|
| Research Abstract |
We developed several statistical methods to estimate causal networks from high-dimensional data and obtain useful causal knowledge. Specifically, we(1) developed two methods to find exogenous variables that trigger causal chains,(2) developed a direct method to estimate the entire or sub-network based on the methods for finding exogenous variables, and(3) evaluated our methods based on simulations on artificial data and real-world datasets including gene-expression data, brain imaging data and sociology data. We further made some software to perform the methods available on the internet so that many practitioners can use our methods.
|
