2016 Fiscal Year Final Research Report
The quantile method for symbolic data analysis.
| Project/Area Number |
25330268
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Tokyo Denki University |
|---|
Principal Investigator |
Ichino Manabu 東京電機大学, 理工学部, 名誉教授 (40057245)
|
|---|
| Research Collaborator |
BRITO Paula Faculty of Economics of the University of Porto, Associate Professor
UMBLEJA Kadri Tallinn University of Technology, Department of Computer Control
|
|---|
| Project Period (FY) |
2013-04-01 – 2017-03-31
|
| Keywords | symbolic data analysis / data mining / Cartesian system model / quantile method / PCA / hierarchical clustering / visualization / regression model |
|---|
| Outline of Final Research Achievements |
We often use intervals and histograms to summarize the given large numbers of numerical data sets. We use the term symbolic data to call such summarized data and data by the aggregation of different data tables.The quantile method transforms the given symbolic data table to a different sized numerical data table by a unified way. Then, we realize various data analysis methods on the transformed data. This research report includes three quantile methods for symbolic data: (1) A hierarchical method of conceptual clustering; (2) Visualization of multidimensional symbolic data; and (3) The lookup table regression model.
|
| Free Research Field |
知能情報学
|
