著者
浦上 大輔 浦田 清 布上 恭子 渡会 雅明 浜野 貢 須田 力 中川 功哉
出版者
Japan Society of Human Growth and Development
雑誌
発育発達研究 (ISSN:13408682)
巻号頁・発行日
vol.1997, no.25, pp.20-28, 1997-07-31 (Released:2010-03-16)
参考文献数
11

The purpose of this study was to investigate the life styles and physical activities of senior high school and technical college students in a snowy region both in snowfall and non-snowfall season. Questionnaires on the life style and physical activity were sent to 1675 students in Hokkaido.The following results were obtained.1) The students went to bed and got up earlier during snowfall season than during non-snowfall season. However, the hours of sleep were almost the same in both seasons.2) More than a half of the students attend to school in snowfall season by different way from non-snowfall season. The commuting time in snowfall season was much longer than that in non-snowfall season.3) More than one-third of the students answered that the amount of their physical activities decreased during the snowfall season. This trend was more noticeable among students who exercised 3 or more times a week than among students who exercised less than 3 times a week.4) For students who did not participate in sports club activities, the shoveling snow in winter was found to be effective to compensate for hypokinetic sensation exercise.5) In snowfall season, students tend to exercise for prolonged time with low frequency and low intensity, however in non-snowfall season they tend to exercise for shorter time with high frequency and high intensity.

2 0 0 0 OA 認知的満足化

著者
高橋 達二 甲野 佑 浦上 大輔
出版者
一般社団法人 人工知能学会
雑誌
人工知能学会論文誌 (ISSN:13460714)
巻号頁・発行日
vol.31, no.6, pp.AI30-M_1-11, 2016-11-01 (Released:2016-12-26)
参考文献数
26
被引用文献数
3

As the scope of reinforcement learning broadens, the number of possible states and of executable actions, and hence the product of the two sets explode. Often, there are more feasible options than allowed trials, because of physical and computational constraints imposed on the agents. In such an occasion, optimization procedures that require first trying all the options once do not work. The situation is what the theory of bounded rationality was proposed to deal with. We formalize the central heuristics of bounded rationality theory named satisficing. Instead of the traditional formulation of satisficing at the policy level in terms of reinforcement learning, we introduce a value function that implements the asymmetric risk attitudes characteristic of human cognition. Operated under the simple greedy policy, the RS (reference satisficing) value function enables an efficient satisficing in K-armed bandit problems, and when the reference level for satisficing is set at an appropriate value, it leads to effective optimization. RS is also tested in a robotic motion learning task in which a robot learns to perform giant-swings (acrobot). While the standard algorithms fail because of the coarse-grained state space, RS shows a stable performance and autonomous exploration that goes without randomized exploration and its gradual annealing necessary for the standard methods.
著者
浦上 大輔 郡司 ペギオ幸夫
出版者
一般社団法人 人工知能学会
雑誌
人工知能学会全国大会論文集 第35回 (2021) (ISSN:27587347)
巻号頁・発行日
pp.1H3GS1b01, 2021 (Released:2021-06-14)

リザバーコンピューティングにおいて時系列データの記憶と分類を担うリザバーは、臨界的な性質を持つことが望ましいと考えられている。しかし、一般的にあるシステムにおいて臨界性を実現するためにはパラメータの微調整を必要とする。一方、我々が提案している非同期セルオートマトン(AT_ECA)はそのようなパラメータ調整を必要とせず、臨界的な時空間パターンを生成する。また、AT_ECAをリザバーとする学習システムは、高い学習能力を有することが明らかになっている。これらを踏まえて、本研究の目的はリザバーの状態を解析するための指標を提案して、AT_ECAの臨界性と学習能力の関係を明らかにすることである。まず初めに、初等セルオートマトン(ECA)をリザバーとした場合について、臨界的な時空間パターンを生成する特定の局所ルールの学習能力が高いことを示し、そのリザバーの状態の特徴を上記の指標によって明らかにする。次に、AT_ECAをリザバーとした場合について、多くの局所ルールで同じ特徴が認められることを示す。これらの結果より、AT_ECAの高い学習能力はその普遍的な臨界性に由来することが明らかになる。