Sparse coding with signal processing on graphs
| Project/Area Number |
16H04362
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Communication/Network engineering
|
|---|
| Research Institution | Tokyo University of Agriculture and Technology |
|---|
Principal Investigator |
Tanaka Yuichi 東京農工大学, 工学(系)研究科(研究院), 准教授 (10547029)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
田中 聡久 東京農工大学, 工学(系)研究科(研究院), 教授 (70360584)
京地 清介 北九州市立大学, 国際環境工学部, 准教授 (70634616)
小野 峻佑 東京工業大学, 科学技術創成研究院, 助教 (60752269)
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥16,900,000 (Direct Cost: ¥13,000,000、Indirect Cost: ¥3,900,000)
Fiscal Year 2018: ¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2017: ¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2016: ¥8,970,000 (Direct Cost: ¥6,900,000、Indirect Cost: ¥2,070,000)
|
|---|
| Keywords | グラフ信号処理 / 機械学習 / 信号処理 / スパース符号化 / センサネットワーク |
|---|
| Outline of Final Research Achievements |
With this grant, we achieved a fast graph sampling method and large-scale singular value thresholding without matrix decomposition.
We employed knowledge of graph signal processing for graph sampling. Particularly, we apply graph localization operators to determine the sampling set. The proposed method is significantly faster than the alternative methods. The signal reconstruction quality of the proposed method also outperformed the existing approaches. For singular value thresholding, we utilized Chebyshev polynomial approximation for the thresholding: We presented that the singular values can be processed without an explicit matrix decomposition.
|
| Academic Significance and Societal Importance of the Research Achievements |
Society 5.0 の実現のためには IoT から得られるビッグデータを効率よく処理する必要がある．本研究の成果は，ビッグデータ解析に対して（グラフ）信号処理の立場から非常に効率的なアルゴリズムを提案したことに学術的意義がある．同時に，今後ますます需要が増すビッグデータを利用した AI に対し，実用的にも大きな一歩となることが期待できる．
|
Report
(4 results)
Research Products
(30 results)
