| Budget Amount *help |
¥96,590,000 (Direct Cost: ¥74,300,000、Indirect Cost: ¥22,290,000)
Fiscal Year 2015: ¥16,510,000 (Direct Cost: ¥12,700,000、Indirect Cost: ¥3,810,000)
Fiscal Year 2014: ¥17,550,000 (Direct Cost: ¥13,500,000、Indirect Cost: ¥4,050,000)
Fiscal Year 2013: ¥20,020,000 (Direct Cost: ¥15,400,000、Indirect Cost: ¥4,620,000)
Fiscal Year 2012: ¥17,680,000 (Direct Cost: ¥13,600,000、Indirect Cost: ¥4,080,000)
Fiscal Year 2011: ¥24,830,000 (Direct Cost: ¥19,100,000、Indirect Cost: ¥5,730,000)
|
|---|
| Outline of Final Research Achievements |
Recently, intelligent data analysis technology based on machine learning has been used in various areas of science and industry. However, even with the state-of-the-art machine learning methods, it is difficult to learn robustly from high-dimensional complex data having strong non-linearity and non-stationarity. In this research project, we developed novel machine learning algorithms for feature selection, feature extraction and reinforcement learning, and demonstrated their usefulness through high-degree-of-freedom humanoid robot control.
|
