カオス的想起を用いた強化学習の高速化とその知的システムへの応用
スポンサーリンク
概要
- 論文の詳細を見る
In an intelligent system such as ITS, the hardware side was strong. In this paper, we paid attention to a software side, that is the human centered system, in the intelligent system. And the display function which was the point of contact of the system with the person was taken up. In this display function, a soft touch display is important. Because, the user becomes a panic when user displays immediately before. However, it is difficult to make such software display beforehand. However, it is difficult to make such software display beforehand. Therefore, we propose the technique which software display knowledge is learned at high speed on-line. This technique is learning method by which Q-Learning is united with Chaotic Evolution. In this paper, this proposed technique is called "Q-Learning with Chaotic Evolution". And the utility is verified by applying to the commuter car system in ITS.
論文 | ランダム
- フォーミングを必要としない抵抗変化メモリ(ReRAM)の作製に関する研究
- 残留磁性による微小磁性物質の非磁性金属塊からの分別検出 : 折れ針の金属ファスナー塊からの分別検出(II)
- 残留磁性による微小磁性物質の非磁性金属塊からの分別検出 : 折れ針の金属ファスナー塊からの分別検出
- 残留磁化による微小磁性物質の非磁性金属塊からの分別検出--折れ針の金属ファスナ-塊からの分別検出(2)
- 残留磁化による微小磁性物質の非磁性金属塊からの分別検出--折れ針の金属ファスナ-塊からの分別検出