| Project/Area Number |
15K00321
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Chubu University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
石井 成郎 一宮研伸大学, 看護学部, 准教授 (80399237)
鈴木 裕利 中部大学, 工学部, 教授 (20340200)
澤野 弘明 愛知工業大学, 情報科学部, 准教授 (10609431)
|
|---|
| Project Period (FY) |
2015-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2017: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2016: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2015: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
|
|---|
| Keywords | 機械学習 / 人工知能 / 協調学習 / 未知の問題領域 / クラウドソーシング / ColBagging / Supervised Actor Critic / 一般回帰ニューラルネットワーク / PSO / スマートグリッド / ブラインドシングルプライスオークション / Game理論 / 人と機械学習による共学習 / 電力卸売り取引 / risk-sensitivek強化学習 / supervised actor-critic / learning on a budget / smart grid / automatic biding system / 教示付きクラウドソーシング |
|---|
| Outline of Final Research Achievements |
Artificial Intelligence generally needs a big data set for its learning. However, in unknown environments, such big data do not exist. To solve this difficulty, we proposed a “collaborative bagging” which is a variation of cloud sourcing technique to make workers solve unknown problems by collaborating with learning machines. For each worker, there is a learning machine that imitates the worker’s behaviors. The solution to candidates from the worker are evaluated according their confidence ratios and translate into a single solution by weighted average. The single solution is used for the feedback-teaching for each worker. As a result, each worker learns the solution to yield better solution candidate in the next time. Therefore, the worker and the learning machine improve their ability gradually by repeating this cycle. Finally, the learning machines become to work as a very excellent problem solver after the latter steps of the learning.
|
| Academic Significance and Societal Importance of the Research Achievements |
実用に供される人工知能は、学習に非常に多くの学習データを必要とする。だが未知の領域では学習データそのものが無いため、人工知能は学習ができない。本研究成果は、この問題を解決する一手段を示した。この手法の最もユニーク且つ重要なポイントは、「人」と「人工知能」の両方の能力を高めるという点である。今後人が、人工知能に頼る場面が増えるならば、人はその能力を失う恐れがある。本手法はこれを防ぐ意味で重要である。
|
