Human Learning Strategy in Multi-Movement Discrimination (Leg Controlled Electric Wheelchair Using an Interactive Learning)
Takeshi Ando, Yasushi KOJIMA, Masatoshi Seki, Kazuya Kawamura, Misato Nihei, Haruhiko Sato, Takenobu Inoue, Masakatsu G. Fujie · TRANSACTIONS OF THE JAPAN SOCIETY OF MECHANICAL ENGINEERS Series C · 2013
Learning is essential for human being. In this article, we analyzed the human learning strategy in the multi-movement discrimination by learning machine. Especially, the leg movements' discrimination using Self Organizing Mapping was taken as an example. Firstly, based on the experiment by ten healthy subjects, learning strategies were divided into two strategies; Convergence strategy (improving the repeatability of each movement) and Independence strategy (conducting different movements). Secondly, a child with severe impairment conducted similar experiments. He selected Convergence strategy and his index values to evaluate the degree of convergence and independence of his leg's movements were quite similar with healthy subjects' value. The generality of Convergence and Independence strategies was suggested in multi-movement discrimination task.