2020 Fiscal Year Final Research Report
Data Mining for Graphs and Networks via Local Intrinsic Dimensional Modeling
Project/Area Number |
18H03296
|
Research Category |
Grant-in-Aid for Scientific Research (B)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Review Section |
Basic Section 61030:Intelligent informatics-related
|
Research Institution | National Institute of Informatics |
Principal Investigator |
Houle Michael 国立情報学研究所, 大学共同利用機関等の部局等, 客員教授 (90399270)
|
Project Period (FY) |
2018-04-01 – 2021-03-31
|
Keywords | 高次元空間 / 極値理論 / デ ー タ マ イ ニ ン グ / 機械学種 / ニ ュ ー ラ ル ネ ッ ト ワ ー ク |
Outline of Final Research Achievements |
The overall aim of this computer science project was to tackle the problem of the curse of dimensionality in similarity applications for complex data types such as feature ensembles, graphs and networks, through the further development of practical techniques that recognize and take advantage of local variations in the intrinsic dimensionality of the data. The main goals were: (1) to advance the existing theory of intrinsic dimensionality to account for combinatorial data types; (2) to confirm the theoretical implications by means of empirical study; (3) to exploit new advances in the theory of local intrinsic dimensionality (LID) to develop more efficient and effective solutions for applications of databases, data mining and multimedia, particularly for graphs, deep neural networks, feature ensembles, and other complex data types.
|
Free Research Field |
data mining, machine learning, similarity search
|
Academic Significance and Societal Importance of the Research Achievements |
この研究プロジェクトは、機械学習とデータマイニング分野のトップ国際会議(ICML、ICMR、KDD、SDM)において、5つの影響力のある出版物を生み出した。この研究は、実際に非常に大きな影響を及ぼし、3年未満で、2つの論文が、合計約500件近く引用されるまでに達した。
|