Development and Application of Singular Model Search Methods Utilizing Singular Regions
Project/Area Number |
25330294
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Soft computing
|
Research Institution | Chubu University |
Principal Investigator |
|
Project Period (FY) |
2013-04-01 – 2016-03-31
|
Project Status |
Completed (Fiscal Year 2015)
|
Budget Amount *help |
¥2,990,000 (Direct Cost: ¥2,300,000、Indirect Cost: ¥690,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2014: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2013: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
|
Keywords | 機械学習 / 特異モデル / ニューラルネット / 探索法 / 特異領域 / 複素ニューラルネット |
Outline of Final Research Achievements |
We have developed a new smart learning method called SSF for a real-valued multilayer perceptron (MLP), which makes good use of singular regions to stably and successively find a series of excellent solutions for MLPs with different hidden units. Moreover, we have developed a new learning method called C-SSF for a complex-valued MLP, which utilizes singular regions and Wirtinger calculus. C-SSF showed excellent performance for both fitting of a heavily swirling function and prediction of chaotic behavior.
|
Report
(4 results)
Research Products
(24 results)