Architectures and optimization algorithms for machine learning from big data
Project/Area Number |
26730114
|
Research Category |
Grant-in-Aid for Young Scientists (B)
|
Allocation Type | Multi-year Fund |
Research Field |
Intelligent informatics
|
Research Institution | The University of Tokyo |
Principal Investigator |
Matsushima Shin 東京大学, 大学院情報理工学系研究科, 常勤講師 (90721837)
|
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 2016: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2015: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2014: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
|
Keywords | 機械学習 / 凸最適化 / スパース学習 / 大規模学習 / SVM / データマイニング / ビッグデータ / 国際情報交換 |
Outline of Final Research Achievements |
In this research, firstly, we proposed an optimization scheme for regularized empirical risk minimization that includes SVM and logistic regression. we have shown that this scheme that performs optimization by operating multiple processes asynchronously allows efficient distributed optimization from both theoretical aspects nd experimental aspect. Secondly, focusing on sparse learning that originally requires several tera-bytes of data, we proposed an optimization scheme that works efficiently by suppressing the amount of data. We have shown that the proposed method can extract features efficiently by using efficient data structure such as suffix array in cases in which substrings are used as features of datasets such as text and DNA.
|
Report
(5 results)
Research Products
(8 results)