Studies on neural basis of prediction error learning theory in insects

2018 Fiscal Year Final Research Report

• Project/Area Number: 16H04814
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Animal physiology/Animal behavior
• Research Institution: Hokkaido University
• Principal Investigator: Mizunami Makoto 北海道大学, 理学研究院, 教授 (30174030)
• Project Period (FY): 2016-04-01 – 2019-03-31
• Keywords: 学習 / 昆虫 / 微小脳 / 報酬 / 罰 / 予測誤差 / オクトパミン / ドーパミン

Outline of Final Research Achievements

I obtained four major findings in this study. At first, I showed that aversive learning in crickets can be accounted for by the prediction error theory. Secondly, I obtained evidence to suggest that dopamine neurons mediate aversive prediction error signals. Thirdly, I was successful to morphologically characterized octopamine-immunoreactive neurons that are likly to convey information about prediction error signals in appetitive learning in cockroaches. Finally, I showed that some features of learning in crickets cannot be accounted for by prediction error theory. In conclusion, I showed applicability of the predictor error to insect learning and also its limitations, and clarified some of its neural basis.

Free Research Field

動物生理・行動学

Academic Significance and Societal Importance of the Research Achievements

本研究は、昆虫の学習の主要な特徴が予測誤差理論で説明できることを初めて明らかにし、またそのニューロン機構の一端を明らかにした画期的なもので、学術的に高い意義がある。
また本研究は、学習の基本アルゴリズムが昆虫と哺乳類の間で共通であることを明らかにし、動物の脳の進化の理解の鍵となる知見をもたらしたもので、社会的にも大きなインパクトを与えうるものである。