Stream data classification with real time learning of kernel matrix
Project/Area Number |
26330251
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Intelligent informatics
|
Research Institution | Prefectural University of Hiroshima (2016) Toyohashi University of Technology (2014-2015) |
Principal Investigator |
Okabe Masayuki 県立広島大学, 経営情報学部, 講師 (50362330)
|
Project Period (FY) |
2014-04-01 – 2017-03-31
|
Project Status |
Completed (Fiscal Year 2016)
|
Budget Amount *help |
¥4,550,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥1,050,000)
Fiscal Year 2016: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2015: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2014: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
|
Keywords | カーネル行列学習 / アンサンブル学習 / 制約付きK-means / 制約付きk-means / ブースティング / 異常検知システム / 異常検知 |
Outline of Final Research Achievements |
Kernel matrix learning is an indispensable technique for machine learning to make of high accuracy. In this research, we developed an algorithm of kernel matrix learning that can be applied to incremental stream data classification and then proposed an active learning method that selects candidate data pairs to be labeled as constraints. We verified the utility of our developed algorithms through the experiments of outlier detection from network traffic.
|
Report
(4 results)
Research Products
(5 results)